id	sid	tid	token	lemma	pos
fuelectenerg-13681	1	1	13681	13681	NUM
fuelectenerg-13681	1	2	facta	facta	PROPN
fuelectenerg-13681	1	3	universitatis	universitatis	PROPN
fuelectenerg-13681	1	4	series	series	NOUN
fuelectenerg-13681	1	5	:	:	PUNCT
fuelectenerg-13681	1	6	electronics	electronic	NOUN
fuelectenerg-13681	1	7	and	and	CCONJ
fuelectenerg-13681	1	8	energetics	energetics	NOUN
fuelectenerg-13681	1	9	vol	vol	NOUN
fuelectenerg-13681	1	10	.	.	PROPN
fuelectenerg-13681	1	11	39	39	NUM
fuelectenerg-13681	1	12	,	,	PUNCT
fuelectenerg-13681	1	13	no	no	DET
fuelectenerg-13681	1	14	1	1	NUM
fuelectenerg-13681	1	15	,	,	PUNCT
fuelectenerg-13681	1	16	march	march	PROPN
fuelectenerg-13681	1	17	2026	2026	NUM
fuelectenerg-13681	1	18	,	,	PUNCT
fuelectenerg-13681	1	19	pp	pp	ADV
fuelectenerg-13681	1	20	.	.	PUNCT
fuelectenerg-13681	1	21	197	197	NUM
fuelectenerg-13681	1	22	217	217	NUM
fuelectenerg-13681	1	23	https://doi.org/10.2298/fuee2601197	https://doi.org/10.2298/fuee2601197	NOUN
fuelectenerg-13681	1	24	m	m	NOUN
fuelectenerg-13681	1	25	©	©	PROPN
fuelectenerg-13681	1	26	2026	2026	NUM
fuelectenerg-13681	1	27	by	by	ADP
fuelectenerg-13681	1	28	university	university	PROPN
fuelectenerg-13681	1	29	of	of	ADP
fuelectenerg-13681	1	30	niš	niš	PROPN
fuelectenerg-13681	1	31	,	,	PUNCT
fuelectenerg-13681	1	32	serbia	serbia	PROPN
fuelectenerg-13681	1	33	|	|	ADV
fuelectenerg-13681	1	34	creative	creative	ADJ
fuelectenerg-13681	1	35	commons	common	NOUN
fuelectenerg-13681	1	36	license	license	NOUN
fuelectenerg-13681	1	37	:	:	PUNCT
fuelectenerg-13681	1	38	cc	cc	VERB
fuelectenerg-13681	1	39	by	by	ADP
fuelectenerg-13681	1	40	-	-	PUNCT
fuelectenerg-13681	1	41	nc	nc	VERB
fuelectenerg-13681	1	42	-	-	PUNCT
fuelectenerg-13681	1	43	nd	nd	ADV
fuelectenerg-13681	1	44	original	original	ADJ
fuelectenerg-13681	1	45	scientific	scientific	ADJ
fuelectenerg-13681	1	46	paper	paper	NOUN
fuelectenerg-13681	1	47	comparative	comparative	ADJ
fuelectenerg-13681	1	48	analysis	analysis	NOUN
fuelectenerg-13681	1	49	of	of	ADP
fuelectenerg-13681	1	50	deep	deep	ADJ
fuelectenerg-13681	1	51	learning	learning	NOUN
fuelectenerg-13681	1	52	and	and	CCONJ
fuelectenerg-13681	1	53	traditional	traditional	ADJ
fuelectenerg-13681	1	54	optimization	optimization	NOUN
fuelectenerg-13681	1	55	algorithms	algorithm	NOUN
fuelectenerg-13681	1	56	for	for	ADP
fuelectenerg-13681	1	57	adaptive	adaptive	ADJ
fuelectenerg-13681	1	58	beamforming	beamforme	VERB
fuelectenerg-13681	1	59	in	in	ADP
fuelectenerg-13681	1	60	mimo	mimo	PROPN
fuelectenerg-13681	1	61	systems	systems	PROPN
fuelectenerg-13681	1	62	mahmoud	mahmoud	PROPN
fuelectenerg-13681	1	63	mohamed	mohamed	PROPN
fuelectenerg-13681	1	64	,	,	PUNCT
fuelectenerg-13681	1	65	fayez	fayez	VERB
fuelectenerg-13681	1	66	aljuaid	aljuaid	PROPN
fuelectenerg-13681	1	67	,	,	PUNCT
fuelectenerg-13681	1	68	mhd	mhd	PROPN
fuelectenerg-13681	1	69	walid	walid	PROPN
fuelectenerg-13681	1	70	koubeisi	koubeisi	VERB
fuelectenerg-13681	1	71	electrical	electrical	ADJ
fuelectenerg-13681	1	72	and	and	CCONJ
fuelectenerg-13681	1	73	computer	computer	NOUN
fuelectenerg-13681	1	74	engineering	engineering	NOUN
fuelectenerg-13681	1	75	,	,	PUNCT
fuelectenerg-13681	1	76	king	king	NOUN
fuelectenerg-13681	1	77	abdul	abdul	PROPN
fuelectenerg-13681	1	78	aziz	aziz	PROPN
fuelectenerg-13681	1	79	university	university	PROPN
fuelectenerg-13681	1	80	,	,	PUNCT
fuelectenerg-13681	1	81	saudi	saudi	PROPN
fuelectenerg-13681	1	82	arabia	arabia	PROPN
fuelectenerg-13681	1	83	orcid	orcid	PROPN
fuelectenerg-13681	1	84	ids	id	NOUN
fuelectenerg-13681	1	85	:	:	PUNCT
fuelectenerg-13681	1	86	mahmoud	mahmoud	PROPN
fuelectenerg-13681	1	87	mohamed	mohamed	PROPN
fuelectenerg-13681	1	88	https://orcid.org/0009-0003-5213-4019	https://orcid.org/0009-0003-5213-4019	PRON
fuelectenerg-13681	1	89	fayez	fayez	VERB
fuelectenerg-13681	1	90	aljuaid	aljuaid	PROPN
fuelectenerg-13681	1	91	https://orcid.org/0009-0008-3889-8983	https://orcid.org/0009-0008-3889-8983	PROPN
fuelectenerg-13681	1	92	mhd	mhd	PROPN
fuelectenerg-13681	1	93	walid	walid	PROPN
fuelectenerg-13681	1	94	koubeisi	koubeisi	PROPN
fuelectenerg-13681	1	95	https://orcid.org/0009-0002-9406-4840	https://orcid.org/0009-0002-9406-4840	NOUN
fuelectenerg-13681	1	96	abstract	abstract	NOUN
fuelectenerg-13681	1	97	.	.	PUNCT
fuelectenerg-13681	2	1	this	this	DET
fuelectenerg-13681	2	2	study	study	NOUN
fuelectenerg-13681	2	3	provides	provide	VERB
fuelectenerg-13681	2	4	a	a	DET
fuelectenerg-13681	2	5	comprehensive	comprehensive	ADJ
fuelectenerg-13681	2	6	methodological	methodological	ADJ
fuelectenerg-13681	2	7	contribution	contribution	NOUN
fuelectenerg-13681	2	8	through	through	ADP
fuelectenerg-13681	2	9	rigorous	rigorous	ADJ
fuelectenerg-13681	2	10	comparative	comparative	ADJ
fuelectenerg-13681	2	11	analysis	analysis	NOUN
fuelectenerg-13681	2	12	between	between	ADP
fuelectenerg-13681	2	13	deep	deep	ADJ
fuelectenerg-13681	2	14	learning	learning	NOUN
fuelectenerg-13681	2	15	approaches	approach	NOUN
fuelectenerg-13681	2	16	and	and	CCONJ
fuelectenerg-13681	2	17	traditional	traditional	ADJ
fuelectenerg-13681	2	18	optimization	optimization	NOUN
fuelectenerg-13681	2	19	algorithms	algorithm	NOUN
fuelectenerg-13681	2	20	for	for	ADP
fuelectenerg-13681	2	21	adaptive	adaptive	ADJ
fuelectenerg-13681	2	22	beamforming	beamforming	NOUN
fuelectenerg-13681	2	23	in	in	ADP
fuelectenerg-13681	2	24	mimo	mimo	PROPN
fuelectenerg-13681	2	25	systems	system	NOUN
fuelectenerg-13681	2	26	.	.	PUNCT
fuelectenerg-13681	3	1	traditional	traditional	ADJ
fuelectenerg-13681	3	2	optimization	optimization	NOUN
fuelectenerg-13681	3	3	methods	method	NOUN
fuelectenerg-13681	3	4	face	face	VERB
fuelectenerg-13681	3	5	significant	significant	ADJ
fuelectenerg-13681	3	6	challenges	challenge	NOUN
fuelectenerg-13681	3	7	in	in	ADP
fuelectenerg-13681	3	8	dynamic	dynamic	ADJ
fuelectenerg-13681	3	9	environments	environment	NOUN
fuelectenerg-13681	3	10	due	due	ADP
fuelectenerg-13681	3	11	to	to	ADP
fuelectenerg-13681	3	12	computational	computational	ADJ
fuelectenerg-13681	3	13	complexity	complexity	NOUN
fuelectenerg-13681	3	14	and	and	CCONJ
fuelectenerg-13681	3	15	convergence	convergence	NOUN
fuelectenerg-13681	3	16	issues	issue	NOUN
fuelectenerg-13681	3	17	.	.	PUNCT
fuelectenerg-13681	4	1	through	through	ADP
fuelectenerg-13681	4	2	systematic	systematic	ADJ
fuelectenerg-13681	4	3	experimentation	experimentation	NOUN
fuelectenerg-13681	4	4	with	with	ADP
fuelectenerg-13681	4	5	standardized	standardized	ADJ
fuelectenerg-13681	4	6	datasets	dataset	NOUN
fuelectenerg-13681	4	7	(	(	PUNCT
fuelectenerg-13681	4	8	deepmimo	deepmimo	VERB
fuelectenerg-13681	4	9	,	,	PUNCT
fuelectenerg-13681	4	10	3gpp	3gpp	PROPN
fuelectenerg-13681	4	11	tr	tr	VERB
fuelectenerg-13681	4	12	38.901	38.901	NUM
fuelectenerg-13681	4	13	,	,	PUNCT
fuelectenerg-13681	4	14	and	and	CCONJ
fuelectenerg-13681	4	15	ieee	ieee	PROPN
fuelectenerg-13681	4	16	mimo	mimo	PROPN
fuelectenerg-13681	4	17	data	data	PROPN
fuelectenerg-13681	4	18	challenge	challenge	PROPN
fuelectenerg-13681	4	19	)	)	PUNCT
fuelectenerg-13681	4	20	,	,	PUNCT
fuelectenerg-13681	4	21	we	we	PRON
fuelectenerg-13681	4	22	evaluate	evaluate	VERB
fuelectenerg-13681	4	23	performance	performance	NOUN
fuelectenerg-13681	4	24	using	use	VERB
fuelectenerg-13681	4	25	statistically	statistically	ADV
fuelectenerg-13681	4	26	validated	validate	VERB
fuelectenerg-13681	4	27	metrics	metric	NOUN
fuelectenerg-13681	4	28	including	include	VERB
fuelectenerg-13681	4	29	signal	signal	NOUN
fuelectenerg-13681	4	30	-	-	PUNCT
fuelectenerg-13681	4	31	to	to	ADP
fuelectenerg-13681	4	32	-	-	PUNCT
fuelectenerg-13681	4	33	interference	interference	NOUN
fuelectenerg-13681	4	34	-	-	PUNCT
fuelectenerg-13681	4	35	plus	plus	CCONJ
fuelectenerg-13681	4	36	-	-	PUNCT
fuelectenerg-13681	4	37	noise	noise	NOUN
fuelectenerg-13681	4	38	ratio	ratio	NOUN
fuelectenerg-13681	4	39	,	,	PUNCT
fuelectenerg-13681	4	40	bit	bit	NOUN
fuelectenerg-13681	4	41	error	error	NOUN
fuelectenerg-13681	4	42	rate	rate	NOUN
fuelectenerg-13681	4	43	,	,	PUNCT
fuelectenerg-13681	4	44	computational	computational	ADJ
fuelectenerg-13681	4	45	efficiency	efficiency	NOUN
fuelectenerg-13681	4	46	,	,	PUNCT
fuelectenerg-13681	4	47	and	and	CCONJ
fuelectenerg-13681	4	48	adaptation	adaptation	NOUN
fuelectenerg-13681	4	49	speed	speed	NOUN
fuelectenerg-13681	4	50	.	.	PUNCT
fuelectenerg-13681	5	1	our	our	PRON
fuelectenerg-13681	5	2	findings	finding	NOUN
fuelectenerg-13681	5	3	reveal	reveal	VERB
fuelectenerg-13681	5	4	that	that	SCONJ
fuelectenerg-13681	5	5	deep	deep	ADJ
fuelectenerg-13681	5	6	learning	learning	NOUN
fuelectenerg-13681	5	7	approaches	approach	NOUN
fuelectenerg-13681	5	8	achieve	achieve	VERB
fuelectenerg-13681	5	9	significantly	significantly	ADV
fuelectenerg-13681	5	10	faster	fast	ADJ
fuelectenerg-13681	5	11	convergence	convergence	NOUN
fuelectenerg-13681	5	12	(	(	PUNCT
fuelectenerg-13681	5	13	23.7	23.7	NUM
fuelectenerg-13681	5	14	%	%	NOUN
fuelectenerg-13681	5	15	,	,	PUNCT
fuelectenerg-13681	5	16	p	p	X
fuelectenerg-13681	5	17	<	<	X
fuelectenerg-13681	5	18	0.01	0.01	NUM
fuelectenerg-13681	5	19	)	)	PUNCT
fuelectenerg-13681	5	20	and	and	CCONJ
fuelectenerg-13681	5	21	higher	high	ADJ
fuelectenerg-13681	5	22	sinr	sinr	NOUN
fuelectenerg-13681	5	23	(	(	PUNCT
fuelectenerg-13681	5	24	18.5	18.5	NUM
fuelectenerg-13681	5	25	%	%	NOUN
fuelectenerg-13681	5	26	,	,	PUNCT
fuelectenerg-13681	5	27	p	p	X
fuelectenerg-13681	5	28	<	<	X
fuelectenerg-13681	5	29	0.01	0.01	NUM
fuelectenerg-13681	5	30	)	)	PUNCT
fuelectenerg-13681	5	31	in	in	ADP
fuelectenerg-13681	5	32	dynamic	dynamic	ADJ
fuelectenerg-13681	5	33	channel	channel	NOUN
fuelectenerg-13681	5	34	conditions	condition	NOUN
fuelectenerg-13681	5	35	,	,	PUNCT
fuelectenerg-13681	5	36	while	while	SCONJ
fuelectenerg-13681	5	37	traditional	traditional	ADJ
fuelectenerg-13681	5	38	algorithms	algorithm	NOUN
fuelectenerg-13681	5	39	maintain	maintain	VERB
fuelectenerg-13681	5	40	superior	superior	ADJ
fuelectenerg-13681	5	41	performance	performance	NOUN
fuelectenerg-13681	5	42	in	in	ADP
fuelectenerg-13681	5	43	steady	steady	ADJ
fuelectenerg-13681	5	44	-	-	PUNCT
fuelectenerg-13681	5	45	state	state	NOUN
fuelectenerg-13681	5	46	scenarios	scenario	NOUN
fuelectenerg-13681	5	47	.	.	PUNCT
fuelectenerg-13681	6	1	traditional	traditional	ADJ
fuelectenerg-13681	6	2	methods	method	NOUN
fuelectenerg-13681	6	3	outperform	outperform	VERB
fuelectenerg-13681	6	4	deep	deep	ADJ
fuelectenerg-13681	6	5	learning	learning	NOUN
fuelectenerg-13681	6	6	by	by	ADP
fuelectenerg-13681	6	7	12.3	12.3	NUM
fuelectenerg-13681	6	8	%	%	NOUN
fuelectenerg-13681	6	9	(	(	PUNCT
fuelectenerg-13681	6	10	p	p	X
fuelectenerg-13681	6	11	<	<	X
fuelectenerg-13681	6	12	0.01	0.01	NUM
fuelectenerg-13681	6	13	)	)	PUNCT
fuelectenerg-13681	6	14	in	in	ADP
fuelectenerg-13681	6	15	terms	term	NOUN
fuelectenerg-13681	6	16	of	of	ADP
fuelectenerg-13681	6	17	ber	ber	NOUN
fuelectenerg-13681	6	18	in	in	ADP
fuelectenerg-13681	6	19	low	low	ADJ
fuelectenerg-13681	6	20	-	-	PUNCT
fuelectenerg-13681	6	21	snr	snr	NOUN
fuelectenerg-13681	6	22	environments	environment	NOUN
fuelectenerg-13681	6	23	.	.	PUNCT
fuelectenerg-13681	7	1	computational	computational	ADJ
fuelectenerg-13681	7	2	complexity	complexity	NOUN
fuelectenerg-13681	7	3	analysis	analysis	NOUN
fuelectenerg-13681	7	4	shows	show	VERB
fuelectenerg-13681	7	5	traditional	traditional	ADJ
fuelectenerg-13681	7	6	methods	method	NOUN
fuelectenerg-13681	7	7	scale	scale	NOUN
fuelectenerg-13681	7	8	as	as	ADP
fuelectenerg-13681	7	9	o(n³	o(n³	PROPN
fuelectenerg-13681	7	10	)	)	PUNCT
fuelectenerg-13681	7	11	with	with	ADP
fuelectenerg-13681	7	12	mimo	mimo	PROPN
fuelectenerg-13681	7	13	size	size	NOUN
fuelectenerg-13681	7	14	n	n	CCONJ
fuelectenerg-13681	7	15	,	,	PUNCT
fuelectenerg-13681	7	16	while	while	SCONJ
fuelectenerg-13681	7	17	deep	deep	ADJ
fuelectenerg-13681	7	18	learning	learning	NOUN
fuelectenerg-13681	7	19	maintains	maintain	VERB
fuelectenerg-13681	7	20	o(n	o(n	NUM
fuelectenerg-13681	7	21	)	)	PUNCT
fuelectenerg-13681	7	22	inference	inference	NOUN
fuelectenerg-13681	7	23	complexity	complexity	NOUN
fuelectenerg-13681	7	24	.	.	PUNCT
fuelectenerg-13681	8	1	the	the	DET
fuelectenerg-13681	8	2	main	main	ADJ
fuelectenerg-13681	8	3	contribution	contribution	NOUN
fuelectenerg-13681	8	4	of	of	ADP
fuelectenerg-13681	8	5	this	this	DET
fuelectenerg-13681	8	6	work	work	NOUN
fuelectenerg-13681	8	7	is	be	AUX
fuelectenerg-13681	8	8	a	a	DET
fuelectenerg-13681	8	9	novel	novel	ADJ
fuelectenerg-13681	8	10	adaptive	adaptive	ADJ
fuelectenerg-13681	8	11	selection	selection	NOUN
fuelectenerg-13681	8	12	framework	framework	NOUN
fuelectenerg-13681	8	13	with	with	ADP
fuelectenerg-13681	8	14	mathematically	mathematically	ADV
fuelectenerg-13681	8	15	proven	prove	VERB
fuelectenerg-13681	8	16	optimality	optimality	NOUN
fuelectenerg-13681	8	17	bounds	bound	NOUN
fuelectenerg-13681	8	18	that	that	PRON
fuelectenerg-13681	8	19	dynamically	dynamically	ADV
fuelectenerg-13681	8	20	switches	switch	VERB
fuelectenerg-13681	8	21	between	between	ADP
fuelectenerg-13681	8	22	methodologies	methodology	NOUN
fuelectenerg-13681	8	23	based	base	VERB
fuelectenerg-13681	8	24	on	on	ADP
fuelectenerg-13681	8	25	current	current	ADJ
fuelectenerg-13681	8	26	channel	channel	NOUN
fuelectenerg-13681	8	27	conditions	condition	NOUN
fuelectenerg-13681	8	28	,	,	PUNCT
fuelectenerg-13681	8	29	achieving	achieve	VERB
fuelectenerg-13681	8	30	15.3	15.3	NUM
fuelectenerg-13681	8	31	%	%	NOUN
fuelectenerg-13681	8	32	higher	high	ADJ
fuelectenerg-13681	8	33	average	average	ADJ
fuelectenerg-13681	8	34	sinr	sinr	NOUN
fuelectenerg-13681	8	35	(	(	PUNCT
fuelectenerg-13681	8	36	p	p	X
fuelectenerg-13681	8	37	<	<	X
fuelectenerg-13681	8	38	0.01	0.01	NUM
fuelectenerg-13681	8	39	)	)	PUNCT
fuelectenerg-13681	8	40	in	in	ADP
fuelectenerg-13681	8	41	mixed	mixed	ADJ
fuelectenerg-13681	8	42	scenarios	scenario	NOUN
fuelectenerg-13681	8	43	compared	compare	VERB
fuelectenerg-13681	8	44	to	to	ADP
fuelectenerg-13681	8	45	fixed	fix	VERB
fuelectenerg-13681	8	46	algorithms	algorithm	NOUN
fuelectenerg-13681	8	47	.	.	PUNCT
fuelectenerg-13681	9	1	key	key	ADJ
fuelectenerg-13681	9	2	words	word	NOUN
fuelectenerg-13681	9	3	:	:	PUNCT
fuelectenerg-13681	9	4	adaptive	adaptive	ADJ
fuelectenerg-13681	9	5	beamforming	beamforming	NOUN
fuelectenerg-13681	9	6	,	,	PUNCT
fuelectenerg-13681	9	7	mimo	mimo	PROPN
fuelectenerg-13681	9	8	systems	systems	PROPN
fuelectenerg-13681	9	9	,	,	PUNCT
fuelectenerg-13681	9	10	deep	deep	ADJ
fuelectenerg-13681	9	11	learning	learning	NOUN
fuelectenerg-13681	9	12	,	,	PUNCT
fuelectenerg-13681	9	13	optimization	optimization	NOUN
fuelectenerg-13681	9	14	algorithms	algorithm	NOUN
fuelectenerg-13681	9	15	,	,	PUNCT
fuelectenerg-13681	9	16	computational	computational	ADJ
fuelectenerg-13681	9	17	complexity	complexity	NOUN
fuelectenerg-13681	9	18	,	,	PUNCT
fuelectenerg-13681	9	19	wireless	wireless	ADJ
fuelectenerg-13681	9	20	communications	communication	NOUN
fuelectenerg-13681	9	21	1	1	NUM
fuelectenerg-13681	9	22	.	.	PUNCT
fuelectenerg-13681	9	23	introduction	introduction	NOUN
fuelectenerg-13681	9	24	the	the	DET
fuelectenerg-13681	9	25	rapid	rapid	ADJ
fuelectenerg-13681	9	26	growth	growth	NOUN
fuelectenerg-13681	9	27	in	in	ADP
fuelectenerg-13681	9	28	wireless	wireless	ADJ
fuelectenerg-13681	9	29	communication	communication	NOUN
fuelectenerg-13681	9	30	demands	demand	NOUN
fuelectenerg-13681	9	31	has	have	AUX
fuelectenerg-13681	9	32	pushed	push	VERB
fuelectenerg-13681	9	33	multiple	multiple	ADJ
fuelectenerg-13681	9	34	-	-	PUNCT
fuelectenerg-13681	9	35	input	input	NOUN
fuelectenerg-13681	9	36	multiple	multiple	ADJ
fuelectenerg-13681	9	37	-	-	PUNCT
fuelectenerg-13681	9	38	output	output	NOUN
fuelectenerg-13681	9	39	(	(	PUNCT
fuelectenerg-13681	9	40	mimo	mimo	PROPN
fuelectenerg-13681	9	41	)	)	PUNCT
fuelectenerg-13681	9	42	systems	system	NOUN
fuelectenerg-13681	9	43	to	to	ADP
fuelectenerg-13681	9	44	the	the	DET
fuelectenerg-13681	9	45	forefront	forefront	NOUN
fuelectenerg-13681	9	46	of	of	ADP
fuelectenerg-13681	9	47	modern	modern	ADJ
fuelectenerg-13681	9	48	communication	communication	NOUN
fuelectenerg-13681	9	49	infrastructure	infrastructure	NOUN
fuelectenerg-13681	9	50	[	[	X
fuelectenerg-13681	9	51	1	1	NUM
fuelectenerg-13681	9	52	,	,	PUNCT
fuelectenerg-13681	9	53	2	2	NUM
fuelectenerg-13681	9	54	]	]	PUNCT
fuelectenerg-13681	9	55	.	.	PUNCT
fuelectenerg-13681	10	1	within	within	ADP
fuelectenerg-13681	10	2	mimo	mimo	PROPN
fuelectenerg-13681	10	3	systems	systems	PROPN
fuelectenerg-13681	10	4	,	,	PUNCT
fuelectenerg-13681	10	5	adaptive	adaptive	ADJ
fuelectenerg-13681	10	6	beamforming	beamforming	NOUN
fuelectenerg-13681	10	7	has	have	AUX
fuelectenerg-13681	10	8	emerged	emerge	VERB
fuelectenerg-13681	10	9	as	as	ADP
fuelectenerg-13681	10	10	a	a	DET
fuelectenerg-13681	10	11	critical	critical	ADJ
fuelectenerg-13681	10	12	technology	technology	NOUN
fuelectenerg-13681	10	13	for	for	SCONJ
fuelectenerg-13681	10	14	received	receive	VERB
fuelectenerg-13681	10	15	may	may	PROPN
fuelectenerg-13681	10	16	13	13	NUM
fuelectenerg-13681	10	17	,	,	PUNCT
fuelectenerg-13681	10	18	2025	2025	NUM
fuelectenerg-13681	10	19	;	;	PUNCT
fuelectenerg-13681	10	20	revised	revise	VERB
fuelectenerg-13681	10	21	june	june	PROPN
fuelectenerg-13681	10	22	23	23	NUM
fuelectenerg-13681	10	23	,	,	PUNCT
fuelectenerg-13681	10	24	2025	2025	NUM
fuelectenerg-13681	10	25	,	,	PUNCT
fuelectenerg-13681	10	26	and	and	CCONJ
fuelectenerg-13681	10	27	august	august	PROPN
fuelectenerg-13681	10	28	08	08	NUM
fuelectenerg-13681	10	29	,	,	PUNCT
fuelectenerg-13681	10	30	2025	2025	NUM
fuelectenerg-13681	10	31	;	;	PUNCT
fuelectenerg-13681	10	32	accepted	accept	VERB
fuelectenerg-13681	10	33	august	august	PROPN
fuelectenerg-13681	10	34	09	09	NUM
fuelectenerg-13681	10	35	,	,	PUNCT
fuelectenerg-13681	10	36	2025	2025	NUM
fuelectenerg-13681	10	37	corresponding	correspond	VERB
fuelectenerg-13681	10	38	author	author	NOUN
fuelectenerg-13681	10	39	:	:	PUNCT
fuelectenerg-13681	10	40	mahmoud	mahmoud	PROPN
fuelectenerg-13681	10	41	mohamed	mohamed	PROPN
fuelectenerg-13681	10	42	electrical	electrical	PROPN
fuelectenerg-13681	10	43	and	and	CCONJ
fuelectenerg-13681	10	44	computer	computer	NOUN
fuelectenerg-13681	10	45	engineering	engineering	NOUN
fuelectenerg-13681	10	46	,	,	PUNCT
fuelectenerg-13681	10	47	king	king	NOUN
fuelectenerg-13681	10	48	abdul	abdul	PROPN
fuelectenerg-13681	10	49	aziz	aziz	PROPN
fuelectenerg-13681	10	50	university	university	PROPN
fuelectenerg-13681	10	51	,	,	PUNCT
fuelectenerg-13681	10	52	saudi	saudi	PROPN
fuelectenerg-13681	10	53	arabia	arabia	PROPN
fuelectenerg-13681	10	54	e	e	PROPN
fuelectenerg-13681	10	55	-	-	NOUN
fuelectenerg-13681	10	56	mail	mail	NOUN
fuelectenerg-13681	10	57	:	:	PUNCT
fuelectenerg-13681	11	1	mhassan0085@stu.kau.edu.sa	mhassan0085@stu.kau.edu.sa	PROPN
fuelectenerg-13681	11	2	mailto:mhassan0085@stu.kau.edu.sa	mailto:mhassan0085@stu.kau.edu.sa	PROPN
fuelectenerg-13681	11	3	198	198	NUM
fuelectenerg-13681	11	4	m.	m.	NOUN
fuelectenerg-13681	11	5	mohamed	mohamed	PROPN
fuelectenerg-13681	11	6	,	,	PUNCT
fuelectenerg-13681	11	7	f.	f.	PROPN
fuelectenerg-13681	11	8	aljuaid	aljuaid	PROPN
fuelectenerg-13681	11	9	,	,	PUNCT
fuelectenerg-13681	11	10	mhd	mhd	PROPN
fuelectenerg-13681	11	11	w.	w.	PROPN
fuelectenerg-13681	11	12	koubeisi	koubeisi	PROPN
fuelectenerg-13681	11	13	enhancing	enhance	VERB
fuelectenerg-13681	11	14	spectral	spectral	ADJ
fuelectenerg-13681	11	15	efficiency	efficiency	NOUN
fuelectenerg-13681	11	16	,	,	PUNCT
fuelectenerg-13681	11	17	signal	signal	ADJ
fuelectenerg-13681	11	18	quality	quality	NOUN
fuelectenerg-13681	11	19	,	,	PUNCT
fuelectenerg-13681	11	20	and	and	CCONJ
fuelectenerg-13681	11	21	network	network	NOUN
fuelectenerg-13681	11	22	capacity	capacity	NOUN
fuelectenerg-13681	11	23	through	through	ADP
fuelectenerg-13681	11	24	intelligent	intelligent	ADJ
fuelectenerg-13681	11	25	spatial	spatial	ADJ
fuelectenerg-13681	11	26	filtering	filtering	NOUN
fuelectenerg-13681	12	1	[	[	X
fuelectenerg-13681	12	2	3	3	NUM
fuelectenerg-13681	12	3	]	]	PUNCT
fuelectenerg-13681	12	4	.	.	PUNCT
fuelectenerg-13681	13	1	this	this	DET
fuelectenerg-13681	13	2	technique	technique	NOUN
fuelectenerg-13681	13	3	strategically	strategically	ADV
fuelectenerg-13681	13	4	optimizes	optimize	VERB
fuelectenerg-13681	13	5	transmission	transmission	NOUN
fuelectenerg-13681	13	6	and	and	CCONJ
fuelectenerg-13681	13	7	reception	reception	NOUN
fuelectenerg-13681	13	8	patterns	pattern	NOUN
fuelectenerg-13681	13	9	to	to	PART
fuelectenerg-13681	13	10	maximize	maximize	VERB
fuelectenerg-13681	13	11	signal	signal	ADJ
fuelectenerg-13681	13	12	power	power	NOUN
fuelectenerg-13681	13	13	toward	toward	ADP
fuelectenerg-13681	13	14	desired	desire	VERB
fuelectenerg-13681	13	15	users	user	NOUN
fuelectenerg-13681	13	16	while	while	SCONJ
fuelectenerg-13681	13	17	minimizing	minimize	VERB
fuelectenerg-13681	13	18	interference	interference	NOUN
fuelectenerg-13681	13	19	toward	toward	ADP
fuelectenerg-13681	13	20	others	other	NOUN
fuelectenerg-13681	13	21	[	[	X
fuelectenerg-13681	13	22	4	4	NUM
fuelectenerg-13681	13	23	]	]	PUNCT
fuelectenerg-13681	13	24	.	.	PUNCT
fuelectenerg-13681	14	1	traditional	traditional	ADJ
fuelectenerg-13681	14	2	optimization	optimization	NOUN
fuelectenerg-13681	14	3	algorithms	algorithm	NOUN
fuelectenerg-13681	14	4	for	for	ADP
fuelectenerg-13681	14	5	adaptive	adaptive	ADJ
fuelectenerg-13681	14	6	beamforming	beamforming	NOUN
fuelectenerg-13681	14	7	,	,	PUNCT
fuelectenerg-13681	14	8	including	include	VERB
fuelectenerg-13681	14	9	least	least	ADJ
fuelectenerg-13681	14	10	mean	mean	ADJ
fuelectenerg-13681	14	11	squares	square	NOUN
fuelectenerg-13681	14	12	(	(	PUNCT
fuelectenerg-13681	14	13	lms	lms	PROPN
fuelectenerg-13681	14	14	)	)	PUNCT
fuelectenerg-13681	14	15	,	,	PUNCT
fuelectenerg-13681	14	16	recursive	recursive	ADJ
fuelectenerg-13681	14	17	least	least	ADJ
fuelectenerg-13681	14	18	squares	square	NOUN
fuelectenerg-13681	14	19	(	(	PUNCT
fuelectenerg-13681	14	20	rls	rls	PROPN
fuelectenerg-13681	14	21	)	)	PUNCT
fuelectenerg-13681	14	22	,	,	PUNCT
fuelectenerg-13681	14	23	and	and	CCONJ
fuelectenerg-13681	14	24	minimum	minimum	ADJ
fuelectenerg-13681	14	25	variance	variance	NOUN
fuelectenerg-13681	14	26	distortionless	distortionless	NOUN
fuelectenerg-13681	14	27	response	response	NOUN
fuelectenerg-13681	14	28	(	(	PUNCT
fuelectenerg-13681	14	29	mvdr	mvdr	NOUN
fuelectenerg-13681	14	30	)	)	PUNCT
fuelectenerg-13681	14	31	,	,	PUNCT
fuelectenerg-13681	14	32	have	have	AUX
fuelectenerg-13681	14	33	been	be	AUX
fuelectenerg-13681	14	34	extensively	extensively	ADV
fuelectenerg-13681	14	35	studied	study	VERB
fuelectenerg-13681	14	36	and	and	CCONJ
fuelectenerg-13681	14	37	implemented	implement	VERB
fuelectenerg-13681	14	38	in	in	ADP
fuelectenerg-13681	14	39	various	various	ADJ
fuelectenerg-13681	14	40	systems	system	NOUN
fuelectenerg-13681	14	41	[	[	X
fuelectenerg-13681	14	42	5	5	NUM
fuelectenerg-13681	14	43	,	,	PUNCT
fuelectenerg-13681	14	44	6	6	NUM
fuelectenerg-13681	14	45	]	]	PUNCT
fuelectenerg-13681	14	46	.	.	PUNCT
fuelectenerg-13681	15	1	these	these	DET
fuelectenerg-13681	15	2	algorithms	algorithm	NOUN
fuelectenerg-13681	15	3	typically	typically	ADV
fuelectenerg-13681	15	4	depend	depend	VERB
fuelectenerg-13681	15	5	on	on	ADP
fuelectenerg-13681	15	6	statistical	statistical	ADJ
fuelectenerg-13681	15	7	properties	property	NOUN
fuelectenerg-13681	15	8	of	of	ADP
fuelectenerg-13681	15	9	signals	signal	NOUN
fuelectenerg-13681	15	10	and	and	CCONJ
fuelectenerg-13681	15	11	mathematical	mathematical	ADJ
fuelectenerg-13681	15	12	optimization	optimization	NOUN
fuelectenerg-13681	15	13	frameworks	framework	NOUN
fuelectenerg-13681	15	14	to	to	PART
fuelectenerg-13681	15	15	iteratively	iteratively	ADV
fuelectenerg-13681	15	16	adjust	adjust	VERB
fuelectenerg-13681	15	17	beamforming	beamforme	VERB
fuelectenerg-13681	15	18	weights	weight	NOUN
fuelectenerg-13681	15	19	.	.	PUNCT
fuelectenerg-13681	16	1	while	while	SCONJ
fuelectenerg-13681	16	2	they	they	PRON
fuelectenerg-13681	16	3	have	have	AUX
fuelectenerg-13681	16	4	proven	prove	VERB
fuelectenerg-13681	16	5	effective	effective	ADJ
fuelectenerg-13681	16	6	in	in	ADP
fuelectenerg-13681	16	7	many	many	ADJ
fuelectenerg-13681	16	8	controlled	control	VERB
fuelectenerg-13681	16	9	environments	environment	NOUN
fuelectenerg-13681	16	10	,	,	PUNCT
fuelectenerg-13681	16	11	these	these	DET
fuelectenerg-13681	16	12	traditional	traditional	ADJ
fuelectenerg-13681	16	13	approaches	approach	NOUN
fuelectenerg-13681	16	14	face	face	VERB
fuelectenerg-13681	16	15	growing	grow	VERB
fuelectenerg-13681	16	16	challenges	challenge	NOUN
fuelectenerg-13681	16	17	in	in	ADP
fuelectenerg-13681	16	18	modern	modern	ADJ
fuelectenerg-13681	16	19	communication	communication	NOUN
fuelectenerg-13681	16	20	scenarios	scenario	NOUN
fuelectenerg-13681	16	21	characterized	characterize	VERB
fuelectenerg-13681	16	22	by	by	ADP
fuelectenerg-13681	16	23	high	high	ADJ
fuelectenerg-13681	16	24	mobility	mobility	NOUN
fuelectenerg-13681	16	25	,	,	PUNCT
fuelectenerg-13681	16	26	dense	dense	ADJ
fuelectenerg-13681	16	27	networks	network	NOUN
fuelectenerg-13681	16	28	,	,	PUNCT
fuelectenerg-13681	16	29	and	and	CCONJ
fuelectenerg-13681	16	30	non	non	ADJ
fuelectenerg-13681	16	31	-	-	ADJ
fuelectenerg-13681	16	32	stationary	stationary	ADJ
fuelectenerg-13681	16	33	channel	channel	NOUN
fuelectenerg-13681	16	34	conditions	condition	NOUN
fuelectenerg-13681	16	35	[	[	X
fuelectenerg-13681	16	36	7	7	NUM
fuelectenerg-13681	16	37	]	]	PUNCT
fuelectenerg-13681	16	38	.	.	PUNCT
fuelectenerg-13681	17	1	more	more	ADV
fuelectenerg-13681	17	2	recently	recently	ADV
fuelectenerg-13681	17	3	,	,	PUNCT
fuelectenerg-13681	17	4	deep	deep	ADJ
fuelectenerg-13681	17	5	learning	learning	NOUN
fuelectenerg-13681	17	6	approaches	approach	NOUN
fuelectenerg-13681	17	7	have	have	AUX
fuelectenerg-13681	17	8	gained	gain	VERB
fuelectenerg-13681	17	9	traction	traction	NOUN
fuelectenerg-13681	17	10	as	as	ADP
fuelectenerg-13681	17	11	promising	promise	VERB
fuelectenerg-13681	17	12	alternatives	alternative	NOUN
fuelectenerg-13681	17	13	for	for	ADP
fuelectenerg-13681	17	14	complex	complex	ADJ
fuelectenerg-13681	17	15	signal	signal	ADJ
fuelectenerg-13681	17	16	processing	processing	NOUN
fuelectenerg-13681	17	17	tasks	task	NOUN
fuelectenerg-13681	17	18	in	in	ADP
fuelectenerg-13681	17	19	wireless	wireless	ADJ
fuelectenerg-13681	17	20	communications	communication	NOUN
fuelectenerg-13681	17	21	[	[	X
fuelectenerg-13681	17	22	8	8	NUM
fuelectenerg-13681	17	23	,	,	PUNCT
fuelectenerg-13681	17	24	9	9	NUM
fuelectenerg-13681	17	25	]	]	PUNCT
fuelectenerg-13681	17	26	.	.	PUNCT
fuelectenerg-13681	18	1	by	by	ADP
fuelectenerg-13681	18	2	leveraging	leverage	VERB
fuelectenerg-13681	18	3	neural	neural	ADJ
fuelectenerg-13681	18	4	networks	network	NOUN
fuelectenerg-13681	18	5	'	'	PART
fuelectenerg-13681	18	6	ability	ability	NOUN
fuelectenerg-13681	18	7	to	to	PART
fuelectenerg-13681	18	8	model	model	VERB
fuelectenerg-13681	18	9	complex	complex	ADJ
fuelectenerg-13681	18	10	nonlinear	nonlinear	ADJ
fuelectenerg-13681	18	11	relationships	relationship	NOUN
fuelectenerg-13681	18	12	and	and	CCONJ
fuelectenerg-13681	18	13	adapt	adapt	VERB
fuelectenerg-13681	18	14	to	to	ADP
fuelectenerg-13681	18	15	changing	change	VERB
fuelectenerg-13681	18	16	environments	environment	NOUN
fuelectenerg-13681	18	17	,	,	PUNCT
fuelectenerg-13681	18	18	deep	deep	ADJ
fuelectenerg-13681	18	19	learning	learning	NOUN
fuelectenerg-13681	18	20	-	-	PUNCT
fuelectenerg-13681	18	21	based	base	VERB
fuelectenerg-13681	18	22	beamforming	beamforming	NOUN
fuelectenerg-13681	18	23	methods	method	NOUN
fuelectenerg-13681	18	24	potentially	potentially	ADV
fuelectenerg-13681	18	25	offer	offer	VERB
fuelectenerg-13681	18	26	solutions	solution	NOUN
fuelectenerg-13681	18	27	to	to	ADP
fuelectenerg-13681	18	28	scenarios	scenario	NOUN
fuelectenerg-13681	18	29	where	where	SCONJ
fuelectenerg-13681	18	30	traditional	traditional	ADJ
fuelectenerg-13681	18	31	methods	method	NOUN
fuelectenerg-13681	18	32	struggle	struggle	VERB
fuelectenerg-13681	18	33	[	[	X
fuelectenerg-13681	18	34	10	10	NUM
fuelectenerg-13681	18	35	]	]	PUNCT
fuelectenerg-13681	18	36	.	.	PUNCT
fuelectenerg-13681	19	1	however	however	ADV
fuelectenerg-13681	19	2	,	,	PUNCT
fuelectenerg-13681	19	3	there	there	PRON
fuelectenerg-13681	19	4	remains	remain	VERB
fuelectenerg-13681	19	5	a	a	DET
fuelectenerg-13681	19	6	gap	gap	NOUN
fuelectenerg-13681	19	7	in	in	ADP
fuelectenerg-13681	19	8	understanding	understanding	NOUN
fuelectenerg-13681	19	9	exactly	exactly	ADV
fuelectenerg-13681	19	10	how	how	SCONJ
fuelectenerg-13681	19	11	these	these	DET
fuelectenerg-13681	19	12	fundamentally	fundamentally	ADV
fuelectenerg-13681	19	13	different	different	ADJ
fuelectenerg-13681	19	14	approaches	approach	NOUN
fuelectenerg-13681	19	15	compare	compare	VERB
fuelectenerg-13681	19	16	across	across	ADP
fuelectenerg-13681	19	17	diverse	diverse	ADJ
fuelectenerg-13681	19	18	operating	operating	NOUN
fuelectenerg-13681	19	19	conditions	condition	NOUN
fuelectenerg-13681	19	20	.	.	PUNCT
fuelectenerg-13681	20	1	1.1	1.1	NUM
fuelectenerg-13681	20	2	.	.	PUNCT
fuelectenerg-13681	21	1	problem	problem	NOUN
fuelectenerg-13681	21	2	statement	statement	NOUN
fuelectenerg-13681	21	3	despite	despite	SCONJ
fuelectenerg-13681	21	4	the	the	DET
fuelectenerg-13681	21	5	theoretical	theoretical	ADJ
fuelectenerg-13681	21	6	advantages	advantage	NOUN
fuelectenerg-13681	21	7	of	of	ADP
fuelectenerg-13681	21	8	both	both	CCONJ
fuelectenerg-13681	21	9	traditional	traditional	ADJ
fuelectenerg-13681	21	10	optimization	optimization	NOUN
fuelectenerg-13681	21	11	algorithms	algorithm	NOUN
fuelectenerg-13681	21	12	and	and	CCONJ
fuelectenerg-13681	21	13	deep	deep	ADJ
fuelectenerg-13681	21	14	learning	learning	NOUN
fuelectenerg-13681	21	15	approaches	approach	NOUN
fuelectenerg-13681	21	16	,	,	PUNCT
fuelectenerg-13681	21	17	there	there	PRON
fuelectenerg-13681	21	18	is	be	VERB
fuelectenerg-13681	21	19	still	still	ADV
fuelectenerg-13681	21	20	limited	limited	ADJ
fuelectenerg-13681	21	21	understanding	understanding	NOUN
fuelectenerg-13681	21	22	of	of	ADP
fuelectenerg-13681	21	23	their	their	PRON
fuelectenerg-13681	21	24	comparative	comparative	ADJ
fuelectenerg-13681	21	25	performance	performance	NOUN
fuelectenerg-13681	21	26	across	across	ADP
fuelectenerg-13681	21	27	the	the	DET
fuelectenerg-13681	21	28	diverse	diverse	ADJ
fuelectenerg-13681	21	29	operating	operating	NOUN
fuelectenerg-13681	21	30	conditions	condition	NOUN
fuelectenerg-13681	21	31	relevant	relevant	ADJ
fuelectenerg-13681	21	32	to	to	ADP
fuelectenerg-13681	21	33	modern	modern	ADJ
fuelectenerg-13681	21	34	mimo	mimo	PROPN
fuelectenerg-13681	21	35	systems	system	NOUN
fuelectenerg-13681	21	36	.	.	PUNCT
fuelectenerg-13681	22	1	this	this	DET
fuelectenerg-13681	22	2	knowledge	knowledge	NOUN
fuelectenerg-13681	22	3	gap	gap	NOUN
fuelectenerg-13681	22	4	makes	make	VERB
fuelectenerg-13681	22	5	it	it	PRON
fuelectenerg-13681	22	6	difficult	difficult	ADJ
fuelectenerg-13681	22	7	for	for	SCONJ
fuelectenerg-13681	22	8	engineers	engineer	NOUN
fuelectenerg-13681	22	9	and	and	CCONJ
fuelectenerg-13681	22	10	researchers	researcher	NOUN
fuelectenerg-13681	22	11	to	to	PART
fuelectenerg-13681	22	12	make	make	VERB
fuelectenerg-13681	22	13	informed	informed	ADJ
fuelectenerg-13681	22	14	decisions	decision	NOUN
fuelectenerg-13681	22	15	when	when	SCONJ
fuelectenerg-13681	22	16	designing	design	VERB
fuelectenerg-13681	22	17	adaptive	adaptive	ADJ
fuelectenerg-13681	22	18	beamforming	beamforming	NOUN
fuelectenerg-13681	22	19	systems	system	NOUN
fuelectenerg-13681	22	20	for	for	ADP
fuelectenerg-13681	22	21	next	next	ADJ
fuelectenerg-13681	22	22	-	-	PUNCT
fuelectenerg-13681	22	23	generation	generation	NOUN
fuelectenerg-13681	22	24	wireless	wireless	NOUN
fuelectenerg-13681	22	25	networks	network	NOUN
fuelectenerg-13681	22	26	.	.	PUNCT
fuelectenerg-13681	23	1	specifically	specifically	ADV
fuelectenerg-13681	23	2	,	,	PUNCT
fuelectenerg-13681	23	3	several	several	ADJ
fuelectenerg-13681	23	4	critical	critical	ADJ
fuelectenerg-13681	23	5	questions	question	NOUN
fuelectenerg-13681	23	6	remain	remain	AUX
fuelectenerg-13681	23	7	inadequately	inadequately	ADV
fuelectenerg-13681	23	8	addressed	address	VERB
fuelectenerg-13681	23	9	:	:	PUNCT
fuelectenerg-13681	24	1	1	1	X
fuelectenerg-13681	24	2	.	.	X
fuelectenerg-13681	24	3	how	how	SCONJ
fuelectenerg-13681	24	4	do	do	AUX
fuelectenerg-13681	24	5	deep	deep	ADJ
fuelectenerg-13681	24	6	learning	learning	NOUN
fuelectenerg-13681	24	7	approaches	approach	NOUN
fuelectenerg-13681	24	8	truly	truly	ADV
fuelectenerg-13681	24	9	compare	compare	VERB
fuelectenerg-13681	24	10	to	to	ADP
fuelectenerg-13681	24	11	traditional	traditional	ADJ
fuelectenerg-13681	24	12	optimization	optimization	NOUN
fuelectenerg-13681	24	13	algorithms	algorithm	NOUN
fuelectenerg-13681	24	14	in	in	ADP
fuelectenerg-13681	24	15	terms	term	NOUN
fuelectenerg-13681	24	16	of	of	ADP
fuelectenerg-13681	24	17	beamforming	beamforme	VERB
fuelectenerg-13681	24	18	accuracy	accuracy	NOUN
fuelectenerg-13681	24	19	,	,	PUNCT
fuelectenerg-13681	24	20	computational	computational	ADJ
fuelectenerg-13681	24	21	efficiency	efficiency	NOUN
fuelectenerg-13681	24	22	,	,	PUNCT
fuelectenerg-13681	24	23	and	and	CCONJ
fuelectenerg-13681	24	24	adaptability	adaptability	NOUN
fuelectenerg-13681	24	25	across	across	ADP
fuelectenerg-13681	24	26	varying	vary	VERB
fuelectenerg-13681	24	27	channel	channel	NOUN
fuelectenerg-13681	24	28	conditions	condition	NOUN
fuelectenerg-13681	24	29	?	?	PUNCT
fuelectenerg-13681	25	1	2	2	X
fuelectenerg-13681	25	2	.	.	X
fuelectenerg-13681	26	1	under	under	ADP
fuelectenerg-13681	26	2	what	what	PRON
fuelectenerg-13681	26	3	specific	specific	ADJ
fuelectenerg-13681	26	4	operating	operating	NOUN
fuelectenerg-13681	26	5	conditions	condition	NOUN
fuelectenerg-13681	26	6	does	do	AUX
fuelectenerg-13681	26	7	each	each	DET
fuelectenerg-13681	26	8	approach	approach	NOUN
fuelectenerg-13681	26	9	demonstrate	demonstrate	VERB
fuelectenerg-13681	26	10	superior	superior	ADJ
fuelectenerg-13681	26	11	performance	performance	NOUN
fuelectenerg-13681	26	12	?	?	PUNCT
fuelectenerg-13681	27	1	3	3	X
fuelectenerg-13681	27	2	.	.	X
fuelectenerg-13681	27	3	is	be	AUX
fuelectenerg-13681	27	4	it	it	PRON
fuelectenerg-13681	27	5	possible	possible	ADJ
fuelectenerg-13681	27	6	to	to	PART
fuelectenerg-13681	27	7	develop	develop	VERB
fuelectenerg-13681	27	8	hybrid	hybrid	ADJ
fuelectenerg-13681	27	9	approaches	approach	NOUN
fuelectenerg-13681	27	10	that	that	PRON
fuelectenerg-13681	27	11	leverage	leverage	VERB
fuelectenerg-13681	27	12	the	the	DET
fuelectenerg-13681	27	13	complementary	complementary	ADJ
fuelectenerg-13681	27	14	strengths	strength	NOUN
fuelectenerg-13681	27	15	of	of	ADP
fuelectenerg-13681	27	16	both	both	DET
fuelectenerg-13681	27	17	methodologies	methodology	NOUN
fuelectenerg-13681	27	18	?	?	PUNCT
fuelectenerg-13681	28	1	research	research	NOUN
fuelectenerg-13681	28	2	gap	gap	NOUN
fuelectenerg-13681	28	3	.	.	PUNCT
fuelectenerg-13681	29	1	current	current	ADJ
fuelectenerg-13681	29	2	literature	literature	NOUN
fuelectenerg-13681	29	3	has	have	AUX
fuelectenerg-13681	29	4	explored	explore	VERB
fuelectenerg-13681	29	5	various	various	ADJ
fuelectenerg-13681	29	6	optimization	optimization	NOUN
fuelectenerg-13681	29	7	techniques	technique	NOUN
fuelectenerg-13681	29	8	[	[	X
fuelectenerg-13681	29	9	11	11	NUM
fuelectenerg-13681	29	10	,	,	PUNCT
fuelectenerg-13681	29	11	12	12	NUM
fuelectenerg-13681	29	12	]	]	PUNCT
fuelectenerg-13681	29	13	and	and	CCONJ
fuelectenerg-13681	29	14	neural	neural	ADJ
fuelectenerg-13681	29	15	network	network	NOUN
fuelectenerg-13681	29	16	architectures	architecture	NOUN
fuelectenerg-13681	29	17	[	[	X
fuelectenerg-13681	29	18	13	13	NUM
fuelectenerg-13681	29	19	,	,	PUNCT
fuelectenerg-13681	29	20	14	14	NUM
fuelectenerg-13681	29	21	]	]	PUNCT
fuelectenerg-13681	29	22	for	for	ADP
fuelectenerg-13681	29	23	beamforming	beamforme	VERB
fuelectenerg-13681	29	24	,	,	PUNCT
fuelectenerg-13681	29	25	but	but	CCONJ
fuelectenerg-13681	29	26	most	most	ADJ
fuelectenerg-13681	29	27	studies	study	NOUN
fuelectenerg-13681	29	28	focus	focus	VERB
fuelectenerg-13681	29	29	on	on	ADP
fuelectenerg-13681	29	30	either	either	CCONJ
fuelectenerg-13681	29	31	traditional	traditional	ADJ
fuelectenerg-13681	29	32	methods	method	NOUN
fuelectenerg-13681	29	33	or	or	CCONJ
fuelectenerg-13681	29	34	deep	deep	ADJ
fuelectenerg-13681	29	35	learning	learning	NOUN
fuelectenerg-13681	29	36	approaches	approach	NOUN
fuelectenerg-13681	29	37	in	in	ADP
fuelectenerg-13681	29	38	isolation	isolation	NOUN
fuelectenerg-13681	29	39	.	.	PUNCT
fuelectenerg-13681	30	1	the	the	DET
fuelectenerg-13681	30	2	handful	handful	NOUN
fuelectenerg-13681	30	3	of	of	ADP
fuelectenerg-13681	30	4	comparative	comparative	ADJ
fuelectenerg-13681	30	5	studies	study	NOUN
fuelectenerg-13681	30	6	available	available	ADJ
fuelectenerg-13681	30	7	[	[	X
fuelectenerg-13681	30	8	15	15	NUM
fuelectenerg-13681	30	9	]	]	PUNCT
fuelectenerg-13681	30	10	typically	typically	ADV
fuelectenerg-13681	30	11	restrict	restrict	VERB
fuelectenerg-13681	30	12	their	their	PRON
fuelectenerg-13681	30	13	analysis	analysis	NOUN
fuelectenerg-13681	30	14	to	to	ADP
fuelectenerg-13681	30	15	specific	specific	ADJ
fuelectenerg-13681	30	16	network	network	NOUN
fuelectenerg-13681	30	17	conditions	condition	NOUN
fuelectenerg-13681	30	18	or	or	CCONJ
fuelectenerg-13681	30	19	limited	limited	ADJ
fuelectenerg-13681	30	20	performance	performance	NOUN
fuelectenerg-13681	30	21	metrics	metric	NOUN
fuelectenerg-13681	30	22	,	,	PUNCT
fuelectenerg-13681	30	23	failing	fail	VERB
fuelectenerg-13681	30	24	to	to	PART
fuelectenerg-13681	30	25	provide	provide	VERB
fuelectenerg-13681	30	26	a	a	DET
fuelectenerg-13681	30	27	comprehensive	comprehensive	ADJ
fuelectenerg-13681	30	28	evaluation	evaluation	NOUN
fuelectenerg-13681	30	29	framework	framework	NOUN
fuelectenerg-13681	30	30	.	.	PUNCT
fuelectenerg-13681	31	1	just	just	ADV
fuelectenerg-13681	31	2	as	as	ADP
fuelectenerg-13681	31	3	problematic	problematic	ADJ
fuelectenerg-13681	31	4	,	,	PUNCT
fuelectenerg-13681	31	5	existing	exist	VERB
fuelectenerg-13681	31	6	research	research	NOUN
fuelectenerg-13681	31	7	often	often	ADV
fuelectenerg-13681	31	8	overlooks	overlook	VERB
fuelectenerg-13681	31	9	practical	practical	ADJ
fuelectenerg-13681	31	10	implementation	implementation	NOUN
fuelectenerg-13681	31	11	aspects	aspect	NOUN
fuelectenerg-13681	31	12	like	like	ADP
fuelectenerg-13681	31	13	computational	computational	ADJ
fuelectenerg-13681	31	14	complexity	complexity	NOUN
fuelectenerg-13681	31	15	,	,	PUNCT
fuelectenerg-13681	31	16	training	training	NOUN
fuelectenerg-13681	31	17	overhead	overhead	NOUN
fuelectenerg-13681	31	18	,	,	PUNCT
fuelectenerg-13681	31	19	and	and	CCONJ
fuelectenerg-13681	31	20	robustness	robustness	NOUN
fuelectenerg-13681	31	21	to	to	ADP
fuelectenerg-13681	31	22	model	model	NOUN
fuelectenerg-13681	31	23	mismatch	mismatch	NOUN
fuelectenerg-13681	31	24	–	–	PUNCT
fuelectenerg-13681	31	25	factors	factor	NOUN
fuelectenerg-13681	31	26	that	that	PRON
fuelectenerg-13681	31	27	prove	prove	VERB
fuelectenerg-13681	31	28	crucial	crucial	ADJ
fuelectenerg-13681	31	29	for	for	ADP
fuelectenerg-13681	31	30	real	real	ADJ
fuelectenerg-13681	31	31	-	-	PUNCT
fuelectenerg-13681	31	32	world	world	NOUN
fuelectenerg-13681	31	33	deployment	deployment	NOUN
fuelectenerg-13681	31	34	[	[	X
fuelectenerg-13681	31	35	16	16	NUM
fuelectenerg-13681	31	36	,	,	PUNCT
fuelectenerg-13681	31	37	17	17	NUM
fuelectenerg-13681	31	38	]	]	PUNCT
fuelectenerg-13681	31	39	.	.	PUNCT
fuelectenerg-13681	32	1	1.2	1.2	NUM
fuelectenerg-13681	32	2	.	.	PUNCT
fuelectenerg-13681	32	3	objectives	objective	VERB
fuelectenerg-13681	32	4	our	our	PRON
fuelectenerg-13681	32	5	study	study	NOUN
fuelectenerg-13681	32	6	aims	aim	VERB
fuelectenerg-13681	32	7	to	to	PART
fuelectenerg-13681	32	8	conduct	conduct	VERB
fuelectenerg-13681	32	9	a	a	DET
fuelectenerg-13681	32	10	thorough	thorough	ADJ
fuelectenerg-13681	32	11	,	,	PUNCT
fuelectenerg-13681	32	12	systematic	systematic	ADJ
fuelectenerg-13681	32	13	comparison	comparison	NOUN
fuelectenerg-13681	32	14	between	between	ADP
fuelectenerg-13681	32	15	deep	deep	ADJ
fuelectenerg-13681	32	16	learning	learning	NOUN
fuelectenerg-13681	32	17	approaches	approach	NOUN
fuelectenerg-13681	32	18	and	and	CCONJ
fuelectenerg-13681	32	19	traditional	traditional	ADJ
fuelectenerg-13681	32	20	optimization	optimization	NOUN
fuelectenerg-13681	32	21	algorithms	algorithm	NOUN
fuelectenerg-13681	32	22	for	for	ADP
fuelectenerg-13681	32	23	adaptive	adaptive	ADJ
fuelectenerg-13681	32	24	beamforming	beamforming	NOUN
fuelectenerg-13681	32	25	in	in	ADP
fuelectenerg-13681	32	26	mimo	mimo	PROPN
fuelectenerg-13681	32	27	systems	system	NOUN
fuelectenerg-13681	32	28	.	.	PUNCT
fuelectenerg-13681	33	1	we	we	PRON
fuelectenerg-13681	33	2	have	have	AUX
fuelectenerg-13681	33	3	structured	structure	VERB
fuelectenerg-13681	33	4	our	our	PRON
fuelectenerg-13681	33	5	research	research	NOUN
fuelectenerg-13681	33	6	around	around	ADP
fuelectenerg-13681	33	7	the	the	DET
fuelectenerg-13681	33	8	following	follow	VERB
fuelectenerg-13681	33	9	specific	specific	ADJ
fuelectenerg-13681	33	10	objectives	objective	NOUN
fuelectenerg-13681	33	11	:	:	PUNCT
fuelectenerg-13681	33	12	deep	deep	ADJ
fuelectenerg-13681	33	13	learning	learning	NOUN
fuelectenerg-13681	33	14	and	and	CCONJ
fuelectenerg-13681	33	15	traditional	traditional	ADJ
fuelectenerg-13681	33	16	algorithms	algorithm	NOUN
fuelectenerg-13681	33	17	for	for	ADP
fuelectenerg-13681	33	18	mimo	mimo	PROPN
fuelectenerg-13681	33	19	beamforming	beamforme	VERB
fuelectenerg-13681	33	20	199	199	NUM
fuelectenerg-13681	33	21	1	1	NUM
fuelectenerg-13681	33	22	.	.	PUNCT
fuelectenerg-13681	33	23	to	to	PART
fuelectenerg-13681	33	24	develop	develop	VERB
fuelectenerg-13681	33	25	a	a	DET
fuelectenerg-13681	33	26	unified	unified	ADJ
fuelectenerg-13681	33	27	evaluation	evaluation	NOUN
fuelectenerg-13681	33	28	framework	framework	NOUN
fuelectenerg-13681	33	29	encompassing	encompass	VERB
fuelectenerg-13681	33	30	multiple	multiple	ADJ
fuelectenerg-13681	33	31	performance	performance	NOUN
fuelectenerg-13681	33	32	metrics	metric	NOUN
fuelectenerg-13681	33	33	relevant	relevant	ADJ
fuelectenerg-13681	33	34	to	to	ADP
fuelectenerg-13681	33	35	practical	practical	ADJ
fuelectenerg-13681	33	36	mimo	mimo	PROPN
fuelectenerg-13681	33	37	systems	system	NOUN
fuelectenerg-13681	33	38	.	.	PUNCT
fuelectenerg-13681	34	1	2	2	X
fuelectenerg-13681	34	2	.	.	PUNCT
fuelectenerg-13681	34	3	to	to	PART
fuelectenerg-13681	34	4	analyze	analyze	VERB
fuelectenerg-13681	34	5	the	the	DET
fuelectenerg-13681	34	6	performance	performance	NOUN
fuelectenerg-13681	34	7	of	of	ADP
fuelectenerg-13681	34	8	both	both	DET
fuelectenerg-13681	34	9	approaches	approach	NOUN
fuelectenerg-13681	34	10	across	across	ADP
fuelectenerg-13681	34	11	diverse	diverse	ADJ
fuelectenerg-13681	34	12	channel	channel	NOUN
fuelectenerg-13681	34	13	conditions	condition	NOUN
fuelectenerg-13681	34	14	,	,	PUNCT
fuelectenerg-13681	34	15	network	network	NOUN
fuelectenerg-13681	34	16	topologies	topology	NOUN
fuelectenerg-13681	34	17	,	,	PUNCT
fuelectenerg-13681	34	18	and	and	CCONJ
fuelectenerg-13681	34	19	user	user	NOUN
fuelectenerg-13681	34	20	mobility	mobility	NOUN
fuelectenerg-13681	34	21	scenarios	scenario	NOUN
fuelectenerg-13681	34	22	.	.	PUNCT
fuelectenerg-13681	35	1	3	3	X
fuelectenerg-13681	35	2	.	.	PUNCT
fuelectenerg-13681	35	3	to	to	PART
fuelectenerg-13681	35	4	characterize	characterize	VERB
fuelectenerg-13681	35	5	the	the	DET
fuelectenerg-13681	35	6	computational	computational	ADJ
fuelectenerg-13681	35	7	complexity	complexity	NOUN
fuelectenerg-13681	35	8	and	and	CCONJ
fuelectenerg-13681	35	9	resource	resource	NOUN
fuelectenerg-13681	35	10	requirements	requirement	NOUN
fuelectenerg-13681	35	11	of	of	ADP
fuelectenerg-13681	35	12	both	both	DET
fuelectenerg-13681	35	13	methods	method	NOUN
fuelectenerg-13681	35	14	.	.	PUNCT
fuelectenerg-13681	36	1	4	4	X
fuelectenerg-13681	36	2	.	.	PUNCT
fuelectenerg-13681	36	3	to	to	PART
fuelectenerg-13681	36	4	identify	identify	VERB
fuelectenerg-13681	36	5	specific	specific	ADJ
fuelectenerg-13681	36	6	scenarios	scenario	NOUN
fuelectenerg-13681	36	7	where	where	SCONJ
fuelectenerg-13681	36	8	each	each	DET
fuelectenerg-13681	36	9	approach	approach	NOUN
fuelectenerg-13681	36	10	demonstrates	demonstrate	VERB
fuelectenerg-13681	36	11	superior	superior	ADJ
fuelectenerg-13681	36	12	performance	performance	NOUN
fuelectenerg-13681	36	13	.	.	PUNCT
fuelectenerg-13681	37	1	5	5	X
fuelectenerg-13681	37	2	.	.	PUNCT
fuelectenerg-13681	37	3	to	to	PART
fuelectenerg-13681	37	4	explore	explore	VERB
fuelectenerg-13681	37	5	potential	potential	ADJ
fuelectenerg-13681	37	6	hybrid	hybrid	ADJ
fuelectenerg-13681	37	7	approaches	approach	NOUN
fuelectenerg-13681	37	8	that	that	PRON
fuelectenerg-13681	37	9	leverage	leverage	VERB
fuelectenerg-13681	37	10	the	the	DET
fuelectenerg-13681	37	11	complementary	complementary	ADJ
fuelectenerg-13681	37	12	strengths	strength	NOUN
fuelectenerg-13681	37	13	of	of	ADP
fuelectenerg-13681	37	14	both	both	DET
fuelectenerg-13681	37	15	methodologies	methodology	NOUN
fuelectenerg-13681	37	16	.	.	PUNCT
fuelectenerg-13681	38	1	key	key	ADJ
fuelectenerg-13681	38	2	contributions	contribution	NOUN
fuelectenerg-13681	38	3	.	.	PUNCT
fuelectenerg-13681	39	1	the	the	DET
fuelectenerg-13681	39	2	key	key	ADJ
fuelectenerg-13681	39	3	contributions	contribution	NOUN
fuelectenerg-13681	39	4	of	of	ADP
fuelectenerg-13681	39	5	this	this	DET
fuelectenerg-13681	39	6	paper	paper	NOUN
fuelectenerg-13681	39	7	include	include	VERB
fuelectenerg-13681	39	8	a	a	DET
fuelectenerg-13681	39	9	mathematically	mathematically	ADV
fuelectenerg-13681	39	10	rigorous	rigorous	ADJ
fuelectenerg-13681	39	11	comparative	comparative	ADJ
fuelectenerg-13681	39	12	analysis	analysis	NOUN
fuelectenerg-13681	39	13	framework	framework	NOUN
fuelectenerg-13681	39	14	with	with	ADP
fuelectenerg-13681	39	15	formal	formal	ADJ
fuelectenerg-13681	39	16	performance	performance	NOUN
fuelectenerg-13681	39	17	bounds	bound	NOUN
fuelectenerg-13681	39	18	that	that	PRON
fuelectenerg-13681	39	19	evaluates	evaluate	VERB
fuelectenerg-13681	39	20	traditional	traditional	ADJ
fuelectenerg-13681	39	21	optimization	optimization	NOUN
fuelectenerg-13681	39	22	algorithms	algorithm	NOUN
fuelectenerg-13681	39	23	and	and	CCONJ
fuelectenerg-13681	39	24	deep	deep	ADJ
fuelectenerg-13681	39	25	learning	learning	NOUN
fuelectenerg-13681	39	26	approaches	approach	NOUN
fuelectenerg-13681	39	27	for	for	ADP
fuelectenerg-13681	39	28	adaptive	adaptive	ADJ
fuelectenerg-13681	39	29	beamforming	beamforme	VERB
fuelectenerg-13681	39	30	across	across	ADP
fuelectenerg-13681	39	31	multiple	multiple	ADJ
fuelectenerg-13681	39	32	metrics	metric	NOUN
fuelectenerg-13681	39	33	and	and	CCONJ
fuelectenerg-13681	39	34	operating	operating	NOUN
fuelectenerg-13681	39	35	conditions	condition	NOUN
fuelectenerg-13681	39	36	.	.	PUNCT
fuelectenerg-13681	40	1	we	we	PRON
fuelectenerg-13681	40	2	utilize	utilize	VERB
fuelectenerg-13681	40	3	standardized	standardized	ADJ
fuelectenerg-13681	40	4	benchmark	benchmark	NOUN
fuelectenerg-13681	40	5	datasets	dataset	NOUN
fuelectenerg-13681	40	6	for	for	ADP
fuelectenerg-13681	40	7	comparing	compare	VERB
fuelectenerg-13681	40	8	beamforming	beamforme	VERB
fuelectenerg-13681	40	9	algorithms	algorithm	NOUN
fuelectenerg-13681	40	10	,	,	PUNCT
fuelectenerg-13681	40	11	including	include	VERB
fuelectenerg-13681	40	12	channel	channel	NOUN
fuelectenerg-13681	40	13	models	model	NOUN
fuelectenerg-13681	40	14	derived	derive	VERB
fuelectenerg-13681	40	15	from	from	ADP
fuelectenerg-13681	40	16	field	field	NOUN
fuelectenerg-13681	40	17	measurements	measurement	NOUN
fuelectenerg-13681	40	18	in	in	ADP
fuelectenerg-13681	40	19	diverse	diverse	ADJ
fuelectenerg-13681	40	20	environments	environment	NOUN
fuelectenerg-13681	40	21	with	with	ADP
fuelectenerg-13681	40	22	statistically	statistically	ADV
fuelectenerg-13681	40	23	validated	validate	VERB
fuelectenerg-13681	40	24	characteristics	characteristic	NOUN
fuelectenerg-13681	40	25	.	.	PUNCT
fuelectenerg-13681	41	1	the	the	DET
fuelectenerg-13681	41	2	paper	paper	NOUN
fuelectenerg-13681	41	3	presents	present	VERB
fuelectenerg-13681	41	4	novel	novel	ADJ
fuelectenerg-13681	41	5	theoretical	theoretical	ADJ
fuelectenerg-13681	41	6	insights	insight	NOUN
fuelectenerg-13681	41	7	into	into	ADP
fuelectenerg-13681	41	8	the	the	DET
fuelectenerg-13681	41	9	fundamental	fundamental	ADJ
fuelectenerg-13681	41	10	limitations	limitation	NOUN
fuelectenerg-13681	41	11	of	of	ADP
fuelectenerg-13681	41	12	both	both	DET
fuelectenerg-13681	41	13	approaches	approach	NOUN
fuelectenerg-13681	41	14	,	,	PUNCT
fuelectenerg-13681	41	15	including	include	VERB
fuelectenerg-13681	41	16	proven	prove	VERB
fuelectenerg-13681	41	17	convergence	convergence	NOUN
fuelectenerg-13681	41	18	properties	property	NOUN
fuelectenerg-13681	41	19	and	and	CCONJ
fuelectenerg-13681	41	20	error	error	NOUN
fuelectenerg-13681	41	21	bounds	bound	NOUN
fuelectenerg-13681	41	22	for	for	ADP
fuelectenerg-13681	41	23	different	different	ADJ
fuelectenerg-13681	41	24	channel	channel	NOUN
fuelectenerg-13681	41	25	conditions	condition	NOUN
fuelectenerg-13681	41	26	.	.	PUNCT
fuelectenerg-13681	42	1	we	we	PRON
fuelectenerg-13681	42	2	introduce	introduce	VERB
fuelectenerg-13681	42	3	a	a	DET
fuelectenerg-13681	42	4	new	new	ADJ
fuelectenerg-13681	42	5	algorithmic	algorithmic	ADJ
fuelectenerg-13681	42	6	framework	framework	NOUN
fuelectenerg-13681	42	7	for	for	ADP
fuelectenerg-13681	42	8	hybrid	hybrid	NOUN
fuelectenerg-13681	42	9	beamforming	beamforming	NOUN
fuelectenerg-13681	42	10	selection	selection	NOUN
fuelectenerg-13681	42	11	with	with	ADP
fuelectenerg-13681	42	12	provable	provable	ADJ
fuelectenerg-13681	42	13	optimality	optimality	NOUN
fuelectenerg-13681	42	14	guarantees	guarantee	NOUN
fuelectenerg-13681	42	15	and	and	CCONJ
fuelectenerg-13681	42	16	complexity	complexity	NOUN
fuelectenerg-13681	42	17	analysis	analysis	NOUN
fuelectenerg-13681	42	18	.	.	PUNCT
fuelectenerg-13681	43	1	finally	finally	ADV
fuelectenerg-13681	43	2	,	,	PUNCT
fuelectenerg-13681	43	3	we	we	PRON
fuelectenerg-13681	43	4	provide	provide	VERB
fuelectenerg-13681	43	5	hardware	hardware	NOUN
fuelectenerg-13681	43	6	-	-	PUNCT
fuelectenerg-13681	43	7	validated	validate	VERB
fuelectenerg-13681	43	8	implementation	implementation	NOUN
fuelectenerg-13681	43	9	and	and	CCONJ
fuelectenerg-13681	43	10	performance	performance	NOUN
fuelectenerg-13681	43	11	measurements	measurement	NOUN
fuelectenerg-13681	43	12	on	on	ADP
fuelectenerg-13681	43	13	commercial	commercial	ADJ
fuelectenerg-13681	43	14	sdr	sdr	PROPN
fuelectenerg-13681	43	15	platforms	platform	NOUN
fuelectenerg-13681	43	16	(	(	PUNCT
fuelectenerg-13681	43	17	usrp	usrp	NOUN
fuelectenerg-13681	43	18	x310	x310	NUM
fuelectenerg-13681	43	19	)	)	PUNCT
fuelectenerg-13681	43	20	that	that	PRON
fuelectenerg-13681	43	21	confirm	confirm	VERB
fuelectenerg-13681	43	22	simulation	simulation	NOUN
fuelectenerg-13681	43	23	results	result	NOUN
fuelectenerg-13681	43	24	.	.	PUNCT
fuelectenerg-13681	44	1	paper	paper	NOUN
fuelectenerg-13681	44	2	organization	organization	NOUN
fuelectenerg-13681	44	3	.	.	PUNCT
fuelectenerg-13681	45	1	the	the	DET
fuelectenerg-13681	45	2	remainder	remainder	NOUN
fuelectenerg-13681	45	3	of	of	ADP
fuelectenerg-13681	45	4	this	this	DET
fuelectenerg-13681	45	5	paper	paper	NOUN
fuelectenerg-13681	45	6	is	be	AUX
fuelectenerg-13681	45	7	organized	organize	VERB
fuelectenerg-13681	45	8	as	as	SCONJ
fuelectenerg-13681	45	9	follows	follow	VERB
fuelectenerg-13681	45	10	:	:	PUNCT
fuelectenerg-13681	45	11	section	section	NOUN
fuelectenerg-13681	45	12	2	2	NUM
fuelectenerg-13681	45	13	provides	provide	VERB
fuelectenerg-13681	45	14	a	a	DET
fuelectenerg-13681	45	15	comprehensive	comprehensive	ADJ
fuelectenerg-13681	45	16	review	review	NOUN
fuelectenerg-13681	45	17	of	of	ADP
fuelectenerg-13681	45	18	related	related	ADJ
fuelectenerg-13681	45	19	work	work	NOUN
fuelectenerg-13681	45	20	on	on	ADP
fuelectenerg-13681	45	21	both	both	CCONJ
fuelectenerg-13681	45	22	traditional	traditional	ADJ
fuelectenerg-13681	45	23	and	and	CCONJ
fuelectenerg-13681	45	24	deep	deep	ADV
fuelectenerg-13681	45	25	learningbased	learningbased	ADJ
fuelectenerg-13681	45	26	adaptive	adaptive	ADJ
fuelectenerg-13681	45	27	beamforming	beamforme	VERB
fuelectenerg-13681	45	28	techniques	technique	NOUN
fuelectenerg-13681	45	29	.	.	PUNCT
fuelectenerg-13681	46	1	section	section	NOUN
fuelectenerg-13681	46	2	3	3	NUM
fuelectenerg-13681	46	3	details	detail	NOUN
fuelectenerg-13681	46	4	our	our	PRON
fuelectenerg-13681	46	5	evaluation	evaluation	NOUN
fuelectenerg-13681	46	6	methodology	methodology	NOUN
fuelectenerg-13681	46	7	,	,	PUNCT
fuelectenerg-13681	46	8	including	include	VERB
fuelectenerg-13681	46	9	algorithm	algorithm	NOUN
fuelectenerg-13681	46	10	implementations	implementation	NOUN
fuelectenerg-13681	46	11	,	,	PUNCT
fuelectenerg-13681	46	12	dataset	dataset	VERB
fuelectenerg-13681	46	13	descriptions	description	NOUN
fuelectenerg-13681	46	14	,	,	PUNCT
fuelectenerg-13681	46	15	and	and	CCONJ
fuelectenerg-13681	46	16	performance	performance	NOUN
fuelectenerg-13681	46	17	metrics	metric	NOUN
fuelectenerg-13681	46	18	.	.	PUNCT
fuelectenerg-13681	47	1	in	in	ADP
fuelectenerg-13681	47	2	section	section	NOUN
fuelectenerg-13681	47	3	4	4	NUM
fuelectenerg-13681	47	4	,	,	PUNCT
fuelectenerg-13681	47	5	we	we	PRON
fuelectenerg-13681	47	6	present	present	VERB
fuelectenerg-13681	47	7	and	and	CCONJ
fuelectenerg-13681	47	8	analyze	analyze	VERB
fuelectenerg-13681	47	9	the	the	DET
fuelectenerg-13681	47	10	comparative	comparative	ADJ
fuelectenerg-13681	47	11	results	result	NOUN
fuelectenerg-13681	47	12	across	across	ADP
fuelectenerg-13681	47	13	various	various	ADJ
fuelectenerg-13681	47	14	operating	operating	NOUN
fuelectenerg-13681	47	15	conditions	condition	NOUN
fuelectenerg-13681	47	16	.	.	PUNCT
fuelectenerg-13681	48	1	section	section	NOUN
fuelectenerg-13681	48	2	5	5	NUM
fuelectenerg-13681	48	3	discusses	discuss	VERB
fuelectenerg-13681	48	4	the	the	DET
fuelectenerg-13681	48	5	implications	implication	NOUN
fuelectenerg-13681	48	6	of	of	ADP
fuelectenerg-13681	48	7	our	our	PRON
fuelectenerg-13681	48	8	findings	finding	NOUN
fuelectenerg-13681	48	9	and	and	CCONJ
fuelectenerg-13681	48	10	proposes	propose	VERB
fuelectenerg-13681	48	11	a	a	DET
fuelectenerg-13681	48	12	hybrid	hybrid	ADJ
fuelectenerg-13681	48	13	approach	approach	NOUN
fuelectenerg-13681	48	14	.	.	PUNCT
fuelectenerg-13681	49	1	finally	finally	ADV
fuelectenerg-13681	49	2	,	,	PUNCT
fuelectenerg-13681	49	3	section	section	NOUN
fuelectenerg-13681	49	4	6	6	NUM
fuelectenerg-13681	49	5	concludes	conclude	VERB
fuelectenerg-13681	49	6	the	the	DET
fuelectenerg-13681	49	7	paper	paper	NOUN
fuelectenerg-13681	49	8	and	and	CCONJ
fuelectenerg-13681	49	9	outlines	outline	VERB
fuelectenerg-13681	49	10	directions	direction	NOUN
fuelectenerg-13681	49	11	for	for	ADP
fuelectenerg-13681	49	12	future	future	ADJ
fuelectenerg-13681	49	13	research	research	NOUN
fuelectenerg-13681	49	14	.	.	PUNCT
fuelectenerg-13681	50	1	2	2	X
fuelectenerg-13681	50	2	.	.	X
fuelectenerg-13681	50	3	related	relate	VERB
fuelectenerg-13681	50	4	work	work	NOUN
fuelectenerg-13681	50	5	adaptive	adaptive	ADJ
fuelectenerg-13681	50	6	beamforming	beamforme	VERB
fuelectenerg-13681	50	7	techniques	technique	NOUN
fuelectenerg-13681	50	8	have	have	AUX
fuelectenerg-13681	50	9	been	be	AUX
fuelectenerg-13681	50	10	extensively	extensively	ADV
fuelectenerg-13681	50	11	studied	study	VERB
fuelectenerg-13681	50	12	in	in	ADP
fuelectenerg-13681	50	13	the	the	DET
fuelectenerg-13681	50	14	mimo	mimo	PROPN
fuelectenerg-13681	50	15	systems	systems	PROPN
fuelectenerg-13681	50	16	context	context	NOUN
fuelectenerg-13681	50	17	.	.	PUNCT
fuelectenerg-13681	51	1	this	this	DET
fuelectenerg-13681	51	2	section	section	NOUN
fuelectenerg-13681	51	3	provides	provide	VERB
fuelectenerg-13681	51	4	a	a	DET
fuelectenerg-13681	51	5	comprehensive	comprehensive	ADJ
fuelectenerg-13681	51	6	review	review	NOUN
fuelectenerg-13681	51	7	of	of	ADP
fuelectenerg-13681	51	8	traditional	traditional	ADJ
fuelectenerg-13681	51	9	optimization	optimization	NOUN
fuelectenerg-13681	51	10	algorithms	algorithm	NOUN
fuelectenerg-13681	51	11	and	and	CCONJ
fuelectenerg-13681	51	12	deep	deep	ADJ
fuelectenerg-13681	51	13	learning	learning	NOUN
fuelectenerg-13681	51	14	approaches	approach	NOUN
fuelectenerg-13681	51	15	,	,	PUNCT
fuelectenerg-13681	51	16	highlighting	highlight	VERB
fuelectenerg-13681	51	17	their	their	PRON
fuelectenerg-13681	51	18	respective	respective	ADJ
fuelectenerg-13681	51	19	strengths	strength	NOUN
fuelectenerg-13681	51	20	and	and	CCONJ
fuelectenerg-13681	51	21	limitations	limitation	NOUN
fuelectenerg-13681	51	22	.	.	PUNCT
fuelectenerg-13681	52	1	2.1	2.1	NUM
fuelectenerg-13681	52	2	.	.	PUNCT
fuelectenerg-13681	52	3	traditional	traditional	ADJ
fuelectenerg-13681	52	4	optimization	optimization	NOUN
fuelectenerg-13681	52	5	algorithms	algorithm	NOUN
fuelectenerg-13681	52	6	for	for	ADP
fuelectenerg-13681	52	7	adaptive	adaptive	ADJ
fuelectenerg-13681	52	8	beamforming	beamforme	VERB
fuelectenerg-13681	52	9	traditional	traditional	ADJ
fuelectenerg-13681	52	10	optimization	optimization	NOUN
fuelectenerg-13681	52	11	-	-	PUNCT
fuelectenerg-13681	52	12	based	base	VERB
fuelectenerg-13681	52	13	approaches	approach	NOUN
fuelectenerg-13681	52	14	for	for	ADP
fuelectenerg-13681	52	15	adaptive	adaptive	ADJ
fuelectenerg-13681	52	16	beamforming	beamforming	NOUN
fuelectenerg-13681	52	17	can	can	AUX
fuelectenerg-13681	52	18	be	be	AUX
fuelectenerg-13681	52	19	broadly	broadly	ADV
fuelectenerg-13681	52	20	categorized	categorize	VERB
fuelectenerg-13681	52	21	into	into	ADP
fuelectenerg-13681	52	22	statistical	statistical	ADJ
fuelectenerg-13681	52	23	and	and	CCONJ
fuelectenerg-13681	52	24	deterministic	deterministic	ADJ
fuelectenerg-13681	52	25	methods	method	NOUN
fuelectenerg-13681	52	26	.	.	PUNCT
fuelectenerg-13681	53	1	2.1.1	2.1.1	X
fuelectenerg-13681	53	2	.	.	PUNCT
fuelectenerg-13681	53	3	statistical	statistical	ADJ
fuelectenerg-13681	53	4	optimization	optimization	NOUN
fuelectenerg-13681	53	5	methods	method	NOUN
fuelectenerg-13681	53	6	statistical	statistical	ADJ
fuelectenerg-13681	53	7	methods	method	NOUN
fuelectenerg-13681	53	8	leverage	leverage	VERB
fuelectenerg-13681	53	9	the	the	DET
fuelectenerg-13681	53	10	statistical	statistical	ADJ
fuelectenerg-13681	53	11	properties	property	NOUN
fuelectenerg-13681	53	12	of	of	ADP
fuelectenerg-13681	53	13	received	receive	VERB
fuelectenerg-13681	53	14	signals	signal	NOUN
fuelectenerg-13681	53	15	to	to	PART
fuelectenerg-13681	53	16	optimize	optimize	VERB
fuelectenerg-13681	53	17	beamforming	beamforme	VERB
fuelectenerg-13681	53	18	weights	weight	NOUN
fuelectenerg-13681	53	19	.	.	PUNCT
fuelectenerg-13681	54	1	liu	liu	PROPN
fuelectenerg-13681	54	2	et	et	PROPN
fuelectenerg-13681	54	3	al	al	PROPN
fuelectenerg-13681	54	4	.	.	PUNCT
fuelectenerg-13681	55	1	[	[	X
fuelectenerg-13681	55	2	18	18	NUM
fuelectenerg-13681	55	3	]	]	PUNCT
fuelectenerg-13681	55	4	presented	present	VERB
fuelectenerg-13681	55	5	a	a	DET
fuelectenerg-13681	55	6	comprehensive	comprehensive	ADJ
fuelectenerg-13681	55	7	analysis	analysis	NOUN
fuelectenerg-13681	55	8	of	of	ADP
fuelectenerg-13681	55	9	minimum	minimum	ADJ
fuelectenerg-13681	55	10	mean	mean	NOUN
fuelectenerg-13681	55	11	square	square	NOUN
fuelectenerg-13681	55	12	error	error	NOUN
fuelectenerg-13681	55	13	(	(	PUNCT
fuelectenerg-13681	55	14	mmse	mmse	ADJ
fuelectenerg-13681	55	15	)	)	PUNCT
fuelectenerg-13681	55	16	beamforming	beamforming	NOUN
fuelectenerg-13681	55	17	,	,	PUNCT
fuelectenerg-13681	55	18	demonstrating	demonstrate	VERB
fuelectenerg-13681	55	19	its	its	PRON
fuelectenerg-13681	55	20	effectiveness	effectiveness	NOUN
fuelectenerg-13681	55	21	in	in	ADP
fuelectenerg-13681	55	22	scenarios	scenario	NOUN
fuelectenerg-13681	55	23	with	with	ADP
fuelectenerg-13681	55	24	accurate	accurate	ADJ
fuelectenerg-13681	55	25	channel	channel	NOUN
fuelectenerg-13681	55	26	state	state	NOUN
fuelectenerg-13681	55	27	information	information	NOUN
fuelectenerg-13681	55	28	(	(	PUNCT
fuelectenerg-13681	55	29	csi	csi	PROPN
fuelectenerg-13681	55	30	)	)	PUNCT
fuelectenerg-13681	55	31	.	.	PUNCT
fuelectenerg-13681	56	1	their	their	PRON
fuelectenerg-13681	56	2	work	work	NOUN
fuelectenerg-13681	56	3	achieved	achieve	VERB
fuelectenerg-13681	56	4	a	a	DET
fuelectenerg-13681	56	5	15	15	NUM
fuelectenerg-13681	56	6	%	%	NOUN
fuelectenerg-13681	56	7	improvement	improvement	NOUN
fuelectenerg-13681	56	8	200	200	NUM
fuelectenerg-13681	56	9	m.	m.	NOUN
fuelectenerg-13681	56	10	mohamed	mohamed	PROPN
fuelectenerg-13681	56	11	,	,	PUNCT
fuelectenerg-13681	56	12	f.	f.	PROPN
fuelectenerg-13681	56	13	aljuaid	aljuaid	PROPN
fuelectenerg-13681	56	14	,	,	PUNCT
fuelectenerg-13681	56	15	mhd	mhd	PROPN
fuelectenerg-13681	56	16	w.	w.	PROPN
fuelectenerg-13681	56	17	koubeisi	koubeisi	VERB
fuelectenerg-13681	56	18	in	in	ADP
fuelectenerg-13681	56	19	sinr	sinr	NOUN
fuelectenerg-13681	56	20	compared	compare	VERB
fuelectenerg-13681	56	21	to	to	ADP
fuelectenerg-13681	56	22	non	non	ADJ
fuelectenerg-13681	56	23	-	-	ADJ
fuelectenerg-13681	56	24	adaptive	adaptive	ADJ
fuelectenerg-13681	56	25	techniques	technique	NOUN
fuelectenerg-13681	56	26	,	,	PUNCT
fuelectenerg-13681	56	27	though	though	SCONJ
fuelectenerg-13681	56	28	they	they	PRON
fuelectenerg-13681	56	29	noted	note	VERB
fuelectenerg-13681	56	30	significant	significant	ADJ
fuelectenerg-13681	56	31	performance	performance	NOUN
fuelectenerg-13681	56	32	degradation	degradation	NOUN
fuelectenerg-13681	56	33	in	in	ADP
fuelectenerg-13681	56	34	rapidly	rapidly	ADV
fuelectenerg-13681	56	35	changing	change	VERB
fuelectenerg-13681	56	36	channel	channel	NOUN
fuelectenerg-13681	56	37	conditions	condition	NOUN
fuelectenerg-13681	56	38	.	.	PUNCT
fuelectenerg-13681	57	1	the	the	DET
fuelectenerg-13681	57	2	least	least	ADV
fuelectenerg-13681	57	3	mean	mean	ADJ
fuelectenerg-13681	57	4	squares	square	NOUN
fuelectenerg-13681	57	5	(	(	PUNCT
fuelectenerg-13681	57	6	lms	lm	NOUN
fuelectenerg-13681	57	7	)	)	PUNCT
fuelectenerg-13681	57	8	algorithm	algorithm	NOUN
fuelectenerg-13681	57	9	and	and	CCONJ
fuelectenerg-13681	57	10	its	its	PRON
fuelectenerg-13681	57	11	variants	variant	NOUN
fuelectenerg-13681	57	12	have	have	AUX
fuelectenerg-13681	57	13	been	be	AUX
fuelectenerg-13681	57	14	widely	widely	ADV
fuelectenerg-13681	57	15	applied	apply	VERB
fuelectenerg-13681	57	16	due	due	ADP
fuelectenerg-13681	57	17	to	to	ADP
fuelectenerg-13681	57	18	their	their	PRON
fuelectenerg-13681	57	19	simplicity	simplicity	NOUN
fuelectenerg-13681	57	20	and	and	CCONJ
fuelectenerg-13681	57	21	robustness	robustness	NOUN
fuelectenerg-13681	57	22	.	.	PUNCT
fuelectenerg-13681	58	1	zhang	zhang	PROPN
fuelectenerg-13681	58	2	et	et	PROPN
fuelectenerg-13681	58	3	al	al	PROPN
fuelectenerg-13681	58	4	.	.	PUNCT
fuelectenerg-13681	59	1	[	[	X
fuelectenerg-13681	59	2	19	19	NUM
fuelectenerg-13681	59	3	]	]	PUNCT
fuelectenerg-13681	59	4	proposed	propose	VERB
fuelectenerg-13681	59	5	an	an	DET
fuelectenerg-13681	59	6	improved	improve	VERB
fuelectenerg-13681	59	7	normalized	normalize	VERB
fuelectenerg-13681	59	8	lms	lm	NOUN
fuelectenerg-13681	59	9	algorithm	algorithm	NOUN
fuelectenerg-13681	59	10	that	that	PRON
fuelectenerg-13681	59	11	achieved	achieve	VERB
fuelectenerg-13681	59	12	faster	fast	ADJ
fuelectenerg-13681	59	13	convergence	convergence	NOUN
fuelectenerg-13681	59	14	in	in	ADP
fuelectenerg-13681	59	15	mobile	mobile	ADJ
fuelectenerg-13681	59	16	scenarios	scenario	NOUN
fuelectenerg-13681	59	17	while	while	SCONJ
fuelectenerg-13681	59	18	maintaining	maintain	VERB
fuelectenerg-13681	59	19	low	low	ADJ
fuelectenerg-13681	59	20	computational	computational	ADJ
fuelectenerg-13681	59	21	complexity	complexity	NOUN
fuelectenerg-13681	59	22	.	.	PUNCT
fuelectenerg-13681	60	1	their	their	PRON
fuelectenerg-13681	60	2	approach	approach	NOUN
fuelectenerg-13681	60	3	demonstrated	demonstrate	VERB
fuelectenerg-13681	60	4	a	a	DET
fuelectenerg-13681	60	5	22	22	NUM
fuelectenerg-13681	60	6	%	%	NOUN
fuelectenerg-13681	60	7	reduction	reduction	NOUN
fuelectenerg-13681	60	8	in	in	ADP
fuelectenerg-13681	60	9	convergence	convergence	NOUN
fuelectenerg-13681	60	10	time	time	NOUN
fuelectenerg-13681	60	11	compared	compare	VERB
fuelectenerg-13681	60	12	to	to	ADP
fuelectenerg-13681	60	13	standard	standard	ADJ
fuelectenerg-13681	60	14	lms	lm	NOUN
fuelectenerg-13681	60	15	,	,	PUNCT
fuelectenerg-13681	60	16	though	though	SCONJ
fuelectenerg-13681	60	17	it	it	PRON
fuelectenerg-13681	60	18	still	still	ADV
fuelectenerg-13681	60	19	required	require	VERB
fuelectenerg-13681	60	20	significant	significant	ADJ
fuelectenerg-13681	60	21	adaptation	adaptation	NOUN
fuelectenerg-13681	60	22	time	time	NOUN
fuelectenerg-13681	60	23	in	in	ADP
fuelectenerg-13681	60	24	highly	highly	ADV
fuelectenerg-13681	60	25	dynamic	dynamic	ADJ
fuelectenerg-13681	60	26	environments	environment	NOUN
fuelectenerg-13681	60	27	.	.	PUNCT
fuelectenerg-13681	61	1	recursive	recursive	ADJ
fuelectenerg-13681	61	2	least	least	ADJ
fuelectenerg-13681	61	3	squares	square	NOUN
fuelectenerg-13681	61	4	(	(	PUNCT
fuelectenerg-13681	61	5	rls	rls	PROPN
fuelectenerg-13681	61	6	)	)	PUNCT
fuelectenerg-13681	61	7	algorithms	algorithm	NOUN
fuelectenerg-13681	61	8	offer	offer	VERB
fuelectenerg-13681	61	9	faster	fast	ADJ
fuelectenerg-13681	61	10	convergence	convergence	NOUN
fuelectenerg-13681	61	11	at	at	ADP
fuelectenerg-13681	61	12	the	the	DET
fuelectenerg-13681	61	13	cost	cost	NOUN
fuelectenerg-13681	61	14	of	of	ADP
fuelectenerg-13681	61	15	increased	increase	VERB
fuelectenerg-13681	61	16	computational	computational	ADJ
fuelectenerg-13681	61	17	complexity	complexity	NOUN
fuelectenerg-13681	61	18	.	.	PUNCT
fuelectenerg-13681	62	1	wang	wang	PROPN
fuelectenerg-13681	62	2	et	et	PROPN
fuelectenerg-13681	62	3	al	al	PROPN
fuelectenerg-13681	62	4	.	.	PUNCT
fuelectenerg-13681	63	1	[	[	X
fuelectenerg-13681	63	2	20	20	NUM
fuelectenerg-13681	63	3	]	]	PUNCT
fuelectenerg-13681	63	4	developed	develop	VERB
fuelectenerg-13681	63	5	a	a	DET
fuelectenerg-13681	63	6	low	low	ADJ
fuelectenerg-13681	63	7	-	-	PUNCT
fuelectenerg-13681	63	8	complexity	complexity	NOUN
fuelectenerg-13681	63	9	variant	variant	NOUN
fuelectenerg-13681	63	10	of	of	ADP
fuelectenerg-13681	63	11	rls	rls	PROPN
fuelectenerg-13681	63	12	specifically	specifically	ADV
fuelectenerg-13681	63	13	tailored	tailor	VERB
fuelectenerg-13681	63	14	for	for	ADP
fuelectenerg-13681	63	15	massive	massive	ADJ
fuelectenerg-13681	63	16	mimo	mimo	PROPN
fuelectenerg-13681	63	17	systems	system	NOUN
fuelectenerg-13681	63	18	.	.	PUNCT
fuelectenerg-13681	64	1	their	their	PRON
fuelectenerg-13681	64	2	method	method	NOUN
fuelectenerg-13681	64	3	achieved	achieve	VERB
fuelectenerg-13681	64	4	near	near	ADV
fuelectenerg-13681	64	5	-	-	PUNCT
fuelectenerg-13681	64	6	optimal	optimal	ADJ
fuelectenerg-13681	64	7	performance	performance	NOUN
fuelectenerg-13681	64	8	with	with	ADP
fuelectenerg-13681	64	9	approximately	approximately	ADV
fuelectenerg-13681	64	10	40	40	NUM
fuelectenerg-13681	64	11	%	%	NOUN
fuelectenerg-13681	64	12	reduction	reduction	NOUN
fuelectenerg-13681	64	13	in	in	ADP
fuelectenerg-13681	64	14	computational	computational	ADJ
fuelectenerg-13681	64	15	requirements	requirement	NOUN
fuelectenerg-13681	64	16	compared	compare	VERB
fuelectenerg-13681	64	17	to	to	ADP
fuelectenerg-13681	64	18	conventional	conventional	ADJ
fuelectenerg-13681	64	19	rls	rl	NOUN
fuelectenerg-13681	64	20	,	,	PUNCT
fuelectenerg-13681	64	21	though	though	SCONJ
fuelectenerg-13681	64	22	the	the	DET
fuelectenerg-13681	64	23	algorithm	algorithm	NOUN
fuelectenerg-13681	64	24	still	still	ADV
fuelectenerg-13681	64	25	faced	face	VERB
fuelectenerg-13681	64	26	challenges	challenge	NOUN
fuelectenerg-13681	64	27	in	in	ADP
fuelectenerg-13681	64	28	scenarios	scenario	NOUN
fuelectenerg-13681	64	29	with	with	ADP
fuelectenerg-13681	64	30	limited	limited	ADJ
fuelectenerg-13681	64	31	training	training	NOUN
fuelectenerg-13681	64	32	data	datum	NOUN
fuelectenerg-13681	64	33	.	.	PUNCT
fuelectenerg-13681	65	1	2.1.2	2.1.2	X
fuelectenerg-13681	65	2	.	.	PUNCT
fuelectenerg-13681	65	3	deterministic	deterministic	ADJ
fuelectenerg-13681	65	4	optimization	optimization	NOUN
fuelectenerg-13681	65	5	methods	method	NOUN
fuelectenerg-13681	65	6	deterministic	deterministic	ADJ
fuelectenerg-13681	65	7	methods	method	NOUN
fuelectenerg-13681	65	8	formulate	formulate	VERB
fuelectenerg-13681	65	9	beamforming	beamforme	VERB
fuelectenerg-13681	65	10	as	as	ADP
fuelectenerg-13681	65	11	constrained	constrain	VERB
fuelectenerg-13681	65	12	optimization	optimization	NOUN
fuelectenerg-13681	65	13	problems	problem	NOUN
fuelectenerg-13681	65	14	with	with	ADP
fuelectenerg-13681	65	15	specific	specific	ADJ
fuelectenerg-13681	65	16	objectives	objective	NOUN
fuelectenerg-13681	65	17	.	.	PUNCT
fuelectenerg-13681	66	1	minimum	minimum	ADJ
fuelectenerg-13681	66	2	variance	variance	NOUN
fuelectenerg-13681	66	3	distortionless	distortionless	NOUN
fuelectenerg-13681	66	4	response	response	NOUN
fuelectenerg-13681	66	5	(	(	PUNCT
fuelectenerg-13681	66	6	mvdr	mvdr	NOUN
fuelectenerg-13681	66	7	)	)	PUNCT
fuelectenerg-13681	66	8	and	and	CCONJ
fuelectenerg-13681	66	9	linearly	linearly	ADV
fuelectenerg-13681	66	10	constrained	constrain	VERB
fuelectenerg-13681	66	11	minimum	minimum	ADJ
fuelectenerg-13681	66	12	variance	variance	NOUN
fuelectenerg-13681	66	13	(	(	PUNCT
fuelectenerg-13681	66	14	lcmv	lcmv	VERB
fuelectenerg-13681	66	15	)	)	PUNCT
fuelectenerg-13681	66	16	are	be	AUX
fuelectenerg-13681	66	17	prominent	prominent	ADJ
fuelectenerg-13681	66	18	examples	example	NOUN
fuelectenerg-13681	66	19	in	in	ADP
fuelectenerg-13681	66	20	this	this	DET
fuelectenerg-13681	66	21	category	category	NOUN
fuelectenerg-13681	66	22	.	.	PUNCT
fuelectenerg-13681	67	1	xia	xia	PROPN
fuelectenerg-13681	67	2	et	et	PROPN
fuelectenerg-13681	67	3	al	al	PROPN
fuelectenerg-13681	67	4	.	.	PUNCT
fuelectenerg-13681	68	1	[	[	X
fuelectenerg-13681	68	2	21	21	NUM
fuelectenerg-13681	68	3	]	]	PUNCT
fuelectenerg-13681	68	4	proposed	propose	VERB
fuelectenerg-13681	68	5	a	a	DET
fuelectenerg-13681	68	6	robust	robust	ADJ
fuelectenerg-13681	68	7	mvdr	mvdr	NOUN
fuelectenerg-13681	68	8	beamforming	beamforme	VERB
fuelectenerg-13681	68	9	technique	technique	NOUN
fuelectenerg-13681	68	10	that	that	PRON
fuelectenerg-13681	68	11	addressed	address	VERB
fuelectenerg-13681	68	12	csi	csi	ADJ
fuelectenerg-13681	68	13	uncertainties	uncertainty	NOUN
fuelectenerg-13681	68	14	through	through	ADP
fuelectenerg-13681	68	15	convex	convex	PROPN
fuelectenerg-13681	68	16	optimization	optimization	NOUN
fuelectenerg-13681	68	17	.	.	PUNCT
fuelectenerg-13681	69	1	their	their	PRON
fuelectenerg-13681	69	2	approach	approach	NOUN
fuelectenerg-13681	69	3	demonstrated	demonstrate	VERB
fuelectenerg-13681	69	4	30	30	NUM
fuelectenerg-13681	69	5	%	%	NOUN
fuelectenerg-13681	69	6	improvement	improvement	NOUN
fuelectenerg-13681	69	7	in	in	ADP
fuelectenerg-13681	69	8	interference	interference	NOUN
fuelectenerg-13681	69	9	rejection	rejection	NOUN
fuelectenerg-13681	69	10	compared	compare	VERB
fuelectenerg-13681	69	11	to	to	ADP
fuelectenerg-13681	69	12	conventional	conventional	ADJ
fuelectenerg-13681	69	13	mvdr	mvdr	NOUN
fuelectenerg-13681	69	14	,	,	PUNCT
fuelectenerg-13681	69	15	particularly	particularly	ADV
fuelectenerg-13681	69	16	in	in	ADP
fuelectenerg-13681	69	17	low	low	ADJ
fuelectenerg-13681	69	18	-	-	PUNCT
fuelectenerg-13681	69	19	snr	snr	NOUN
fuelectenerg-13681	69	20	regimes	regime	NOUN
fuelectenerg-13681	69	21	.	.	PUNCT
fuelectenerg-13681	70	1	convex	convex	PROPN
fuelectenerg-13681	70	2	optimization	optimization	NOUN
fuelectenerg-13681	70	3	techniques	technique	NOUN
fuelectenerg-13681	70	4	have	have	AUX
fuelectenerg-13681	70	5	gained	gain	VERB
fuelectenerg-13681	70	6	significant	significant	ADJ
fuelectenerg-13681	70	7	attention	attention	NOUN
fuelectenerg-13681	70	8	due	due	ADP
fuelectenerg-13681	70	9	to	to	ADP
fuelectenerg-13681	70	10	their	their	PRON
fuelectenerg-13681	70	11	theoretical	theoretical	ADJ
fuelectenerg-13681	70	12	guarantees	guarantee	NOUN
fuelectenerg-13681	70	13	.	.	PUNCT
fuelectenerg-13681	71	1	wu	wu	PROPN
fuelectenerg-13681	71	2	et	et	PROPN
fuelectenerg-13681	71	3	al	al	PROPN
fuelectenerg-13681	71	4	.	.	PUNCT
fuelectenerg-13681	72	1	[	[	X
fuelectenerg-13681	72	2	22	22	NUM
fuelectenerg-13681	72	3	]	]	PUNCT
fuelectenerg-13681	72	4	formulated	formulate	VERB
fuelectenerg-13681	72	5	the	the	DET
fuelectenerg-13681	72	6	beamforming	beamforming	NOUN
fuelectenerg-13681	72	7	problem	problem	NOUN
fuelectenerg-13681	72	8	as	as	ADP
fuelectenerg-13681	72	9	a	a	DET
fuelectenerg-13681	72	10	secondorder	secondorder	ADJ
fuelectenerg-13681	72	11	cone	cone	NOUN
fuelectenerg-13681	72	12	program	program	NOUN
fuelectenerg-13681	72	13	(	(	PUNCT
fuelectenerg-13681	72	14	socp	socp	ADJ
fuelectenerg-13681	72	15	)	)	PUNCT
fuelectenerg-13681	72	16	and	and	CCONJ
fuelectenerg-13681	72	17	developed	develop	VERB
fuelectenerg-13681	72	18	an	an	DET
fuelectenerg-13681	72	19	efficient	efficient	ADJ
fuelectenerg-13681	72	20	solution	solution	NOUN
fuelectenerg-13681	72	21	method	method	NOUN
fuelectenerg-13681	72	22	.	.	PUNCT
fuelectenerg-13681	73	1	their	their	PRON
fuelectenerg-13681	73	2	approach	approach	NOUN
fuelectenerg-13681	73	3	achieved	achieve	VERB
fuelectenerg-13681	73	4	near	near	ADP
fuelectenerg-13681	73	5	-	-	PUNCT
fuelectenerg-13681	73	6	optimal	optimal	ADJ
fuelectenerg-13681	73	7	performance	performance	NOUN
fuelectenerg-13681	73	8	with	with	ADP
fuelectenerg-13681	73	9	polynomial	polynomial	ADJ
fuelectenerg-13681	73	10	-	-	PUNCT
fuelectenerg-13681	73	11	time	time	NOUN
fuelectenerg-13681	73	12	complexity	complexity	NOUN
fuelectenerg-13681	73	13	,	,	PUNCT
fuelectenerg-13681	73	14	though	though	SCONJ
fuelectenerg-13681	73	15	its	its	PRON
fuelectenerg-13681	73	16	realtime	realtime	ADJ
fuelectenerg-13681	73	17	implementation	implementation	NOUN
fuelectenerg-13681	73	18	remained	remain	VERB
fuelectenerg-13681	73	19	challenging	challenge	VERB
fuelectenerg-13681	73	20	for	for	ADP
fuelectenerg-13681	73	21	large	large	ADJ
fuelectenerg-13681	73	22	-	-	PUNCT
fuelectenerg-13681	73	23	scale	scale	NOUN
fuelectenerg-13681	73	24	mimo	mimo	PROPN
fuelectenerg-13681	73	25	systems	systems	PROPN
fuelectenerg-13681	73	26	.	.	PUNCT
fuelectenerg-13681	74	1	semidefinite	semidefinite	PROPN
fuelectenerg-13681	74	2	relaxation	relaxation	PROPN
fuelectenerg-13681	74	3	(	(	PUNCT
fuelectenerg-13681	74	4	sdr	sdr	NOUN
fuelectenerg-13681	74	5	)	)	PUNCT
fuelectenerg-13681	74	6	approaches	approach	VERB
fuelectenerg-13681	74	7	convert	convert	VERB
fuelectenerg-13681	74	8	non	non	ADJ
fuelectenerg-13681	74	9	-	-	ADJ
fuelectenerg-13681	74	10	convex	convex	ADJ
fuelectenerg-13681	74	11	beamforming	beamforming	NOUN
fuelectenerg-13681	74	12	problems	problem	NOUN
fuelectenerg-13681	74	13	into	into	ADP
fuelectenerg-13681	74	14	tractable	tractable	ADJ
fuelectenerg-13681	74	15	convex	convex	NOUN
fuelectenerg-13681	74	16	forms	form	NOUN
fuelectenerg-13681	74	17	.	.	PUNCT
fuelectenerg-13681	75	1	guo	guo	PROPN
fuelectenerg-13681	75	2	et	et	PROPN
fuelectenerg-13681	75	3	al	al	PROPN
fuelectenerg-13681	75	4	.	.	PUNCT
fuelectenerg-13681	76	1	[	[	X
fuelectenerg-13681	76	2	23	23	NUM
fuelectenerg-13681	76	3	]	]	PUNCT
fuelectenerg-13681	76	4	applied	apply	VERB
fuelectenerg-13681	76	5	sdr	sdr	PROPN
fuelectenerg-13681	76	6	to	to	ADP
fuelectenerg-13681	76	7	multiuser	multiuser	NOUN
fuelectenerg-13681	76	8	beamforming	beamforme	VERB
fuelectenerg-13681	76	9	and	and	CCONJ
fuelectenerg-13681	76	10	demonstrated	demonstrate	VERB
fuelectenerg-13681	76	11	that	that	SCONJ
fuelectenerg-13681	76	12	the	the	DET
fuelectenerg-13681	76	13	approach	approach	NOUN
fuelectenerg-13681	76	14	achieves	achieve	VERB
fuelectenerg-13681	76	15	performance	performance	NOUN
fuelectenerg-13681	76	16	within	within	ADP
fuelectenerg-13681	76	17	5	5	NUM
fuelectenerg-13681	76	18	%	%	NOUN
fuelectenerg-13681	76	19	of	of	ADP
fuelectenerg-13681	76	20	the	the	DET
fuelectenerg-13681	76	21	global	global	ADJ
fuelectenerg-13681	76	22	optimum	optimum	NOUN
fuelectenerg-13681	76	23	.	.	PUNCT
fuelectenerg-13681	77	1	however	however	ADV
fuelectenerg-13681	77	2	,	,	PUNCT
fuelectenerg-13681	77	3	they	they	PRON
fuelectenerg-13681	77	4	noted	note	VERB
fuelectenerg-13681	77	5	that	that	SCONJ
fuelectenerg-13681	77	6	the	the	DET
fuelectenerg-13681	77	7	computational	computational	ADJ
fuelectenerg-13681	77	8	complexity	complexity	NOUN
fuelectenerg-13681	77	9	scales	scale	VERB
fuelectenerg-13681	77	10	poorly	poorly	ADV
fuelectenerg-13681	77	11	with	with	ADP
fuelectenerg-13681	77	12	the	the	DET
fuelectenerg-13681	77	13	number	number	NOUN
fuelectenerg-13681	77	14	of	of	ADP
fuelectenerg-13681	77	15	users	user	NOUN
fuelectenerg-13681	77	16	and	and	CCONJ
fuelectenerg-13681	77	17	antennas	antenna	NOUN
fuelectenerg-13681	77	18	.	.	PUNCT
fuelectenerg-13681	78	1	2.2	2.2	NUM
fuelectenerg-13681	78	2	.	.	PUNCT
fuelectenerg-13681	79	1	deep	deep	ADJ
fuelectenerg-13681	79	2	learning	learning	NOUN
fuelectenerg-13681	79	3	approaches	approach	NOUN
fuelectenerg-13681	79	4	for	for	ADP
fuelectenerg-13681	79	5	adaptive	adaptive	ADJ
fuelectenerg-13681	79	6	beamforming	beamforme	VERB
fuelectenerg-13681	79	7	recent	recent	ADJ
fuelectenerg-13681	79	8	years	year	NOUN
fuelectenerg-13681	79	9	have	have	AUX
fuelectenerg-13681	79	10	witnessed	witness	VERB
fuelectenerg-13681	79	11	growing	grow	VERB
fuelectenerg-13681	79	12	interest	interest	NOUN
fuelectenerg-13681	79	13	in	in	ADP
fuelectenerg-13681	79	14	applying	apply	VERB
fuelectenerg-13681	79	15	deep	deep	ADJ
fuelectenerg-13681	79	16	learning	learning	NOUN
fuelectenerg-13681	79	17	techniques	technique	NOUN
fuelectenerg-13681	79	18	to	to	ADP
fuelectenerg-13681	79	19	adaptive	adaptive	ADJ
fuelectenerg-13681	79	20	beamforming	beamforming	NOUN
fuelectenerg-13681	79	21	problems	problem	NOUN
fuelectenerg-13681	79	22	,	,	PUNCT
fuelectenerg-13681	79	23	leveraging	leverage	VERB
fuelectenerg-13681	79	24	neural	neural	ADJ
fuelectenerg-13681	79	25	networks	network	NOUN
fuelectenerg-13681	79	26	'	'	PART
fuelectenerg-13681	79	27	ability	ability	NOUN
fuelectenerg-13681	79	28	to	to	PART
fuelectenerg-13681	79	29	model	model	VERB
fuelectenerg-13681	79	30	complex	complex	ADJ
fuelectenerg-13681	79	31	non	non	ADJ
fuelectenerg-13681	79	32	-	-	ADJ
fuelectenerg-13681	79	33	linear	linear	ADJ
fuelectenerg-13681	79	34	relationships	relationship	NOUN
fuelectenerg-13681	79	35	.	.	PUNCT
fuelectenerg-13681	80	1	2.2.1	2.2.1	NUM
fuelectenerg-13681	80	2	.	.	PUNCT
fuelectenerg-13681	80	3	supervised	supervised	ADJ
fuelectenerg-13681	80	4	learning	learning	NOUN
fuelectenerg-13681	80	5	approaches	approach	NOUN
fuelectenerg-13681	80	6	supervised	supervise	VERB
fuelectenerg-13681	80	7	learning	learn	VERB
fuelectenerg-13681	80	8	approaches	approach	NOUN
fuelectenerg-13681	80	9	utilize	utilize	VERB
fuelectenerg-13681	80	10	labeled	label	VERB
fuelectenerg-13681	80	11	training	training	NOUN
fuelectenerg-13681	80	12	data	datum	NOUN
fuelectenerg-13681	80	13	to	to	PART
fuelectenerg-13681	80	14	learn	learn	VERB
fuelectenerg-13681	80	15	the	the	DET
fuelectenerg-13681	80	16	mapping	mapping	NOUN
fuelectenerg-13681	80	17	between	between	ADP
fuelectenerg-13681	80	18	channel	channel	NOUN
fuelectenerg-13681	80	19	conditions	condition	NOUN
fuelectenerg-13681	80	20	and	and	CCONJ
fuelectenerg-13681	80	21	optimal	optimal	ADJ
fuelectenerg-13681	80	22	beamforming	beamforming	NOUN
fuelectenerg-13681	80	23	weights	weight	NOUN
fuelectenerg-13681	80	24	.	.	PUNCT
fuelectenerg-13681	81	1	elbir	elbir	NOUN
fuelectenerg-13681	81	2	and	and	CCONJ
fuelectenerg-13681	81	3	mishra	mishra	PROPN
fuelectenerg-13681	82	1	[	[	X
fuelectenerg-13681	82	2	24	24	NUM
fuelectenerg-13681	82	3	]	]	PUNCT
fuelectenerg-13681	82	4	proposed	propose	VERB
fuelectenerg-13681	82	5	a	a	DET
fuelectenerg-13681	82	6	deep	deep	ADJ
fuelectenerg-13681	82	7	neural	neural	ADJ
fuelectenerg-13681	82	8	network	network	NOUN
fuelectenerg-13681	82	9	(	(	PUNCT
fuelectenerg-13681	82	10	dnn	dnn	PROPN
fuelectenerg-13681	82	11	)	)	PUNCT
fuelectenerg-13681	82	12	architecture	architecture	NOUN
fuelectenerg-13681	82	13	that	that	PRON
fuelectenerg-13681	82	14	directly	directly	ADV
fuelectenerg-13681	82	15	predicted	predict	VERB
fuelectenerg-13681	82	16	beamforming	beamforme	VERB
fuelectenerg-13681	82	17	vectors	vector	NOUN
fuelectenerg-13681	82	18	from	from	ADP
fuelectenerg-13681	82	19	channel	channel	NOUN
fuelectenerg-13681	82	20	matrices	matrix	NOUN
fuelectenerg-13681	82	21	.	.	PUNCT
fuelectenerg-13681	83	1	their	their	PRON
fuelectenerg-13681	83	2	approach	approach	NOUN
fuelectenerg-13681	83	3	achieved	achieve	VERB
fuelectenerg-13681	83	4	performance	performance	NOUN
fuelectenerg-13681	83	5	comparable	comparable	ADJ
fuelectenerg-13681	83	6	to	to	ADP
fuelectenerg-13681	83	7	mmse	mmse	ADJ
fuelectenerg-13681	83	8	beamforming	beamforming	NOUN
fuelectenerg-13681	83	9	with	with	ADP
fuelectenerg-13681	83	10	85	85	NUM
fuelectenerg-13681	83	11	%	%	NOUN
fuelectenerg-13681	83	12	reduction	reduction	NOUN
fuelectenerg-13681	83	13	in	in	ADP
fuelectenerg-13681	83	14	computational	computational	ADJ
fuelectenerg-13681	83	15	complexity	complexity	NOUN
fuelectenerg-13681	83	16	during	during	ADP
fuelectenerg-13681	83	17	inference	inference	NOUN
fuelectenerg-13681	83	18	,	,	PUNCT
fuelectenerg-13681	83	19	though	though	SCONJ
fuelectenerg-13681	83	20	it	it	PRON
fuelectenerg-13681	83	21	required	require	VERB
fuelectenerg-13681	83	22	extensive	extensive	ADJ
fuelectenerg-13681	83	23	offline	offline	ADJ
fuelectenerg-13681	83	24	training	training	NOUN
fuelectenerg-13681	83	25	.	.	PUNCT
fuelectenerg-13681	84	1	deep	deep	ADJ
fuelectenerg-13681	84	2	learning	learning	NOUN
fuelectenerg-13681	84	3	and	and	CCONJ
fuelectenerg-13681	84	4	traditional	traditional	ADJ
fuelectenerg-13681	84	5	algorithms	algorithm	NOUN
fuelectenerg-13681	84	6	for	for	ADP
fuelectenerg-13681	84	7	mimo	mimo	PROPN
fuelectenerg-13681	84	8	beamforming	beamforme	VERB
fuelectenerg-13681	84	9	201	201	NUM
fuelectenerg-13681	84	10	convolutional	convolutional	ADJ
fuelectenerg-13681	84	11	neural	neural	ADJ
fuelectenerg-13681	84	12	networks	network	NOUN
fuelectenerg-13681	84	13	(	(	PUNCT
fuelectenerg-13681	84	14	cnns	cnns	PROPN
fuelectenerg-13681	84	15	)	)	PUNCT
fuelectenerg-13681	84	16	have	have	AUX
fuelectenerg-13681	84	17	been	be	AUX
fuelectenerg-13681	84	18	applied	apply	VERB
fuelectenerg-13681	84	19	to	to	PART
fuelectenerg-13681	84	20	exploit	exploit	VERB
fuelectenerg-13681	84	21	the	the	DET
fuelectenerg-13681	84	22	spatial	spatial	ADJ
fuelectenerg-13681	84	23	structure	structure	NOUN
fuelectenerg-13681	84	24	in	in	ADP
fuelectenerg-13681	84	25	channel	channel	NOUN
fuelectenerg-13681	84	26	matrices	matrix	NOUN
fuelectenerg-13681	84	27	.	.	PUNCT
fuelectenerg-13681	85	1	jin	jin	NOUN
fuelectenerg-13681	85	2	et	et	PROPN
fuelectenerg-13681	85	3	al	al	PROPN
fuelectenerg-13681	85	4	.	.	PUNCT
fuelectenerg-13681	86	1	[	[	X
fuelectenerg-13681	86	2	25	25	NUM
fuelectenerg-13681	86	3	]	]	PUNCT
fuelectenerg-13681	86	4	developed	develop	VERB
fuelectenerg-13681	86	5	a	a	DET
fuelectenerg-13681	86	6	cnn	cnn	PROPN
fuelectenerg-13681	86	7	-	-	PUNCT
fuelectenerg-13681	86	8	based	base	VERB
fuelectenerg-13681	86	9	beamforming	beamforming	NOUN
fuelectenerg-13681	86	10	predictor	predictor	NOUN
fuelectenerg-13681	86	11	that	that	PRON
fuelectenerg-13681	86	12	achieved	achieve	VERB
fuelectenerg-13681	86	13	18	18	NUM
fuelectenerg-13681	86	14	%	%	NOUN
fuelectenerg-13681	86	15	higher	high	ADJ
fuelectenerg-13681	86	16	spectral	spectral	ADJ
fuelectenerg-13681	86	17	efficiency	efficiency	NOUN
fuelectenerg-13681	86	18	compared	compare	VERB
fuelectenerg-13681	86	19	to	to	ADP
fuelectenerg-13681	86	20	conventional	conventional	ADJ
fuelectenerg-13681	86	21	algorithms	algorithm	NOUN
fuelectenerg-13681	86	22	in	in	ADP
fuelectenerg-13681	86	23	scenarios	scenario	NOUN
fuelectenerg-13681	86	24	with	with	ADP
fuelectenerg-13681	86	25	partial	partial	ADJ
fuelectenerg-13681	86	26	csi	csi	NOUN
fuelectenerg-13681	86	27	.	.	PUNCT
fuelectenerg-13681	87	1	their	their	PRON
fuelectenerg-13681	87	2	method	method	NOUN
fuelectenerg-13681	87	3	demonstrated	demonstrate	VERB
fuelectenerg-13681	87	4	particular	particular	ADJ
fuelectenerg-13681	87	5	strength	strength	NOUN
fuelectenerg-13681	87	6	in	in	ADP
fuelectenerg-13681	87	7	environments	environment	NOUN
fuelectenerg-13681	87	8	with	with	ADP
fuelectenerg-13681	87	9	spatial	spatial	ADJ
fuelectenerg-13681	87	10	correlation	correlation	NOUN
fuelectenerg-13681	87	11	,	,	PUNCT
fuelectenerg-13681	87	12	though	though	SCONJ
fuelectenerg-13681	87	13	performance	performance	NOUN
fuelectenerg-13681	87	14	degraded	degrade	VERB
fuelectenerg-13681	87	15	in	in	ADP
fuelectenerg-13681	87	16	completely	completely	ADV
fuelectenerg-13681	87	17	random	random	ADJ
fuelectenerg-13681	87	18	channels	channel	NOUN
fuelectenerg-13681	87	19	.	.	PUNCT
fuelectenerg-13681	88	1	recurrent	recurrent	ADJ
fuelectenerg-13681	88	2	neural	neural	ADJ
fuelectenerg-13681	88	3	networks	network	NOUN
fuelectenerg-13681	88	4	(	(	PUNCT
fuelectenerg-13681	88	5	rnns	rnns	NOUN
fuelectenerg-13681	88	6	)	)	PUNCT
fuelectenerg-13681	88	7	leverage	leverage	NOUN
fuelectenerg-13681	88	8	temporal	temporal	ADJ
fuelectenerg-13681	88	9	correlations	correlation	NOUN
fuelectenerg-13681	88	10	in	in	ADP
fuelectenerg-13681	88	11	channel	channel	NOUN
fuelectenerg-13681	88	12	evolution	evolution	NOUN
fuelectenerg-13681	88	13	.	.	PUNCT
fuelectenerg-13681	89	1	liu	liu	PROPN
fuelectenerg-13681	89	2	et	et	PROPN
fuelectenerg-13681	89	3	al	al	PROPN
fuelectenerg-13681	89	4	.	.	PUNCT
fuelectenerg-13681	90	1	[	[	X
fuelectenerg-13681	90	2	26	26	NUM
fuelectenerg-13681	90	3	]	]	PUNCT
fuelectenerg-13681	90	4	implemented	implement	VERB
fuelectenerg-13681	90	5	a	a	DET
fuelectenerg-13681	90	6	long	long	ADJ
fuelectenerg-13681	90	7	short	short	ADJ
fuelectenerg-13681	90	8	-	-	PUNCT
fuelectenerg-13681	90	9	term	term	NOUN
fuelectenerg-13681	90	10	memory	memory	NOUN
fuelectenerg-13681	90	11	(	(	PUNCT
fuelectenerg-13681	90	12	lstm	lstm	NOUN
fuelectenerg-13681	90	13	)	)	PUNCT
fuelectenerg-13681	90	14	network	network	NOUN
fuelectenerg-13681	90	15	for	for	ADP
fuelectenerg-13681	90	16	predictive	predictive	ADJ
fuelectenerg-13681	90	17	beamforming	beamforming	NOUN
fuelectenerg-13681	90	18	in	in	ADP
fuelectenerg-13681	90	19	mobile	mobile	ADJ
fuelectenerg-13681	90	20	scenarios	scenario	NOUN
fuelectenerg-13681	90	21	.	.	PUNCT
fuelectenerg-13681	91	1	their	their	PRON
fuelectenerg-13681	91	2	approach	approach	NOUN
fuelectenerg-13681	91	3	demonstrated	demonstrate	VERB
fuelectenerg-13681	91	4	25	25	NUM
fuelectenerg-13681	91	5	%	%	NOUN
fuelectenerg-13681	91	6	lower	low	ADJ
fuelectenerg-13681	91	7	bit	bit	NOUN
fuelectenerg-13681	91	8	error	error	NOUN
fuelectenerg-13681	91	9	rate	rate	NOUN
fuelectenerg-13681	91	10	compared	compare	VERB
fuelectenerg-13681	91	11	to	to	ADP
fuelectenerg-13681	91	12	traditional	traditional	ADJ
fuelectenerg-13681	91	13	methods	method	NOUN
fuelectenerg-13681	91	14	in	in	ADP
fuelectenerg-13681	91	15	high	high	ADJ
fuelectenerg-13681	91	16	-	-	PUNCT
fuelectenerg-13681	91	17	mobility	mobility	NOUN
fuelectenerg-13681	91	18	environments	environment	NOUN
fuelectenerg-13681	91	19	by	by	ADP
fuelectenerg-13681	91	20	anticipating	anticipate	VERB
fuelectenerg-13681	91	21	channel	channel	NOUN
fuelectenerg-13681	91	22	variations	variation	NOUN
fuelectenerg-13681	91	23	.	.	PUNCT
fuelectenerg-13681	92	1	2.2.2	2.2.2	X
fuelectenerg-13681	92	2	.	.	PUNCT
fuelectenerg-13681	92	3	unsupervised	unsupervised	ADJ
fuelectenerg-13681	92	4	and	and	CCONJ
fuelectenerg-13681	92	5	reinforcement	reinforcement	NOUN
fuelectenerg-13681	92	6	learning	learning	NOUN
fuelectenerg-13681	92	7	approaches	approach	NOUN
fuelectenerg-13681	92	8	unsupervised	unsupervised	ADJ
fuelectenerg-13681	92	9	and	and	CCONJ
fuelectenerg-13681	92	10	reinforcement	reinforcement	NOUN
fuelectenerg-13681	92	11	learning	learning	NOUN
fuelectenerg-13681	92	12	methods	method	NOUN
fuelectenerg-13681	92	13	reduce	reduce	VERB
fuelectenerg-13681	92	14	or	or	CCONJ
fuelectenerg-13681	92	15	eliminate	eliminate	VERB
fuelectenerg-13681	92	16	the	the	DET
fuelectenerg-13681	92	17	need	need	NOUN
fuelectenerg-13681	92	18	for	for	ADP
fuelectenerg-13681	92	19	labeled	label	VERB
fuelectenerg-13681	92	20	training	training	NOUN
fuelectenerg-13681	92	21	data	datum	NOUN
fuelectenerg-13681	92	22	.	.	PUNCT
fuelectenerg-13681	93	1	gao	gao	PROPN
fuelectenerg-13681	93	2	et	et	PROPN
fuelectenerg-13681	93	3	al	al	PROPN
fuelectenerg-13681	93	4	.	.	PUNCT
fuelectenerg-13681	94	1	[	[	X
fuelectenerg-13681	94	2	28	28	NUM
fuelectenerg-13681	94	3	]	]	PUNCT
fuelectenerg-13681	94	4	proposed	propose	VERB
fuelectenerg-13681	94	5	an	an	DET
fuelectenerg-13681	94	6	autoencoder	autoencoder	NOUN
fuelectenerg-13681	94	7	-	-	PUNCT
fuelectenerg-13681	94	8	based	base	VERB
fuelectenerg-13681	94	9	approach	approach	NOUN
fuelectenerg-13681	94	10	that	that	PRON
fuelectenerg-13681	94	11	learned	learn	VERB
fuelectenerg-13681	94	12	efficient	efficient	ADJ
fuelectenerg-13681	94	13	beamforming	beamforming	NOUN
fuelectenerg-13681	94	14	representations	representation	NOUN
fuelectenerg-13681	94	15	without	without	ADP
fuelectenerg-13681	94	16	explicit	explicit	ADJ
fuelectenerg-13681	94	17	channel	channel	NOUN
fuelectenerg-13681	94	18	models	model	NOUN
fuelectenerg-13681	94	19	.	.	PUNCT
fuelectenerg-13681	95	1	their	their	PRON
fuelectenerg-13681	95	2	method	method	NOUN
fuelectenerg-13681	95	3	demonstrated	demonstrate	VERB
fuelectenerg-13681	95	4	robust	robust	ADJ
fuelectenerg-13681	95	5	performance	performance	NOUN
fuelectenerg-13681	95	6	across	across	ADP
fuelectenerg-13681	95	7	diverse	diverse	ADJ
fuelectenerg-13681	95	8	channel	channel	NOUN
fuelectenerg-13681	95	9	conditions	condition	NOUN
fuelectenerg-13681	95	10	but	but	CCONJ
fuelectenerg-13681	95	11	required	require	VERB
fuelectenerg-13681	95	12	significant	significant	ADJ
fuelectenerg-13681	95	13	fine	fine	NOUN
fuelectenerg-13681	95	14	-	-	PUNCT
fuelectenerg-13681	95	15	tuning	tuning	NOUN
fuelectenerg-13681	95	16	to	to	PART
fuelectenerg-13681	95	17	match	match	VERB
fuelectenerg-13681	95	18	the	the	DET
fuelectenerg-13681	95	19	performance	performance	NOUN
fuelectenerg-13681	95	20	of	of	ADP
fuelectenerg-13681	95	21	supervised	supervised	ADJ
fuelectenerg-13681	95	22	approaches	approach	NOUN
fuelectenerg-13681	95	23	.	.	PUNCT
fuelectenerg-13681	96	1	reinforcement	reinforcement	NOUN
fuelectenerg-13681	96	2	learning	learning	NOUN
fuelectenerg-13681	96	3	(	(	PUNCT
fuelectenerg-13681	96	4	rl	rl	NOUN
fuelectenerg-13681	96	5	)	)	PUNCT
fuelectenerg-13681	96	6	formulates	formulate	VERB
fuelectenerg-13681	96	7	beamforming	beamforme	VERB
fuelectenerg-13681	96	8	as	as	ADP
fuelectenerg-13681	96	9	a	a	DET
fuelectenerg-13681	96	10	sequential	sequential	ADJ
fuelectenerg-13681	96	11	decisionmaking	decisionmaking	NOUN
fuelectenerg-13681	96	12	problem	problem	NOUN
fuelectenerg-13681	96	13	.	.	PUNCT
fuelectenerg-13681	97	1	chaudhary	chaudhary	PROPN
fuelectenerg-13681	97	2	et	et	PROPN
fuelectenerg-13681	97	3	al	al	PROPN
fuelectenerg-13681	97	4	.	.	PUNCT
fuelectenerg-13681	98	1	[	[	X
fuelectenerg-13681	98	2	19	19	NUM
fuelectenerg-13681	98	3	]	]	PUNCT
fuelectenerg-13681	98	4	developed	develop	VERB
fuelectenerg-13681	98	5	a	a	DET
fuelectenerg-13681	98	6	deep	deep	ADJ
fuelectenerg-13681	98	7	q	q	NOUN
fuelectenerg-13681	98	8	-	-	PUNCT
fuelectenerg-13681	98	9	network	network	NOUN
fuelectenerg-13681	98	10	(	(	PUNCT
fuelectenerg-13681	98	11	dqn	dqn	PROPN
fuelectenerg-13681	98	12	)	)	PUNCT
fuelectenerg-13681	98	13	for	for	ADP
fuelectenerg-13681	98	14	dynamic	dynamic	ADJ
fuelectenerg-13681	98	15	beamforming	beamforming	NOUN
fuelectenerg-13681	98	16	that	that	PRON
fuelectenerg-13681	98	17	continuously	continuously	ADV
fuelectenerg-13681	98	18	adapted	adapt	VERB
fuelectenerg-13681	98	19	to	to	ADP
fuelectenerg-13681	98	20	changing	change	VERB
fuelectenerg-13681	98	21	interference	interference	NOUN
fuelectenerg-13681	98	22	patterns	pattern	NOUN
fuelectenerg-13681	98	23	.	.	PUNCT
fuelectenerg-13681	99	1	their	their	PRON
fuelectenerg-13681	99	2	approach	approach	NOUN
fuelectenerg-13681	99	3	outperformed	outperform	VERB
fuelectenerg-13681	99	4	traditional	traditional	ADJ
fuelectenerg-13681	99	5	algorithms	algorithm	NOUN
fuelectenerg-13681	99	6	by	by	ADP
fuelectenerg-13681	99	7	12	12	NUM
fuelectenerg-13681	99	8	%	%	NOUN
fuelectenerg-13681	99	9	in	in	ADP
fuelectenerg-13681	99	10	terms	term	NOUN
fuelectenerg-13681	99	11	of	of	ADP
fuelectenerg-13681	99	12	average	average	ADJ
fuelectenerg-13681	99	13	sinr	sinr	NOUN
fuelectenerg-13681	99	14	in	in	ADP
fuelectenerg-13681	99	15	interferencelimited	interferencelimite	VERB
fuelectenerg-13681	99	16	scenarios	scenario	NOUN
fuelectenerg-13681	99	17	,	,	PUNCT
fuelectenerg-13681	99	18	though	though	SCONJ
fuelectenerg-13681	99	19	convergence	convergence	NOUN
fuelectenerg-13681	99	20	time	time	NOUN
fuelectenerg-13681	99	21	remained	remain	VERB
fuelectenerg-13681	99	22	a	a	DET
fuelectenerg-13681	99	23	challenge	challenge	NOUN
fuelectenerg-13681	99	24	.	.	PUNCT
fuelectenerg-13681	100	1	model	model	NOUN
fuelectenerg-13681	100	2	-	-	PUNCT
fuelectenerg-13681	100	3	based	base	VERB
fuelectenerg-13681	100	4	deep	deep	ADJ
fuelectenerg-13681	100	5	learning	learning	NOUN
fuelectenerg-13681	100	6	combines	combine	VERB
fuelectenerg-13681	100	7	traditional	traditional	ADJ
fuelectenerg-13681	100	8	signal	signal	NOUN
fuelectenerg-13681	100	9	processing	processing	NOUN
fuelectenerg-13681	100	10	with	with	ADP
fuelectenerg-13681	100	11	neural	neural	ADJ
fuelectenerg-13681	100	12	networks	network	NOUN
fuelectenerg-13681	100	13	.	.	PUNCT
fuelectenerg-13681	101	1	xia	xia	PROPN
fuelectenerg-13681	101	2	et	et	PROPN
fuelectenerg-13681	101	3	al	al	PROPN
fuelectenerg-13681	101	4	.	.	PUNCT
fuelectenerg-13681	102	1	[	[	X
fuelectenerg-13681	102	2	27	27	NUM
fuelectenerg-13681	102	3	]	]	PUNCT
fuelectenerg-13681	102	4	proposed	propose	VERB
fuelectenerg-13681	102	5	an	an	DET
fuelectenerg-13681	102	6	unfolded	unfolded	ADJ
fuelectenerg-13681	102	7	iterative	iterative	NOUN
fuelectenerg-13681	102	8	algorithm	algorithm	NOUN
fuelectenerg-13681	102	9	where	where	SCONJ
fuelectenerg-13681	102	10	a	a	DET
fuelectenerg-13681	102	11	neural	neural	ADJ
fuelectenerg-13681	102	12	network	network	NOUN
fuelectenerg-13681	102	13	learned	learn	VERB
fuelectenerg-13681	102	14	to	to	PART
fuelectenerg-13681	102	15	optimize	optimize	VERB
fuelectenerg-13681	102	16	each	each	DET
fuelectenerg-13681	102	17	iteration	iteration	NOUN
fuelectenerg-13681	102	18	parameters	parameter	NOUN
fuelectenerg-13681	102	19	.	.	PUNCT
fuelectenerg-13681	103	1	this	this	DET
fuelectenerg-13681	103	2	approach	approach	NOUN
fuelectenerg-13681	103	3	achieved	achieve	VERB
fuelectenerg-13681	103	4	35	35	NUM
fuelectenerg-13681	103	5	%	%	NOUN
fuelectenerg-13681	103	6	faster	fast	ADJ
fuelectenerg-13681	103	7	convergence	convergence	NOUN
fuelectenerg-13681	103	8	than	than	ADP
fuelectenerg-13681	103	9	purely	purely	ADV
fuelectenerg-13681	103	10	analytical	analytical	ADJ
fuelectenerg-13681	103	11	methods	method	NOUN
fuelectenerg-13681	103	12	while	while	SCONJ
fuelectenerg-13681	103	13	maintaining	maintain	VERB
fuelectenerg-13681	103	14	interpretability	interpretability	NOUN
fuelectenerg-13681	103	15	.	.	PUNCT
fuelectenerg-13681	104	1	2.3	2.3	NUM
fuelectenerg-13681	104	2	.	.	PUNCT
fuelectenerg-13681	104	3	hybrid	hybrid	ADJ
fuelectenerg-13681	104	4	and	and	CCONJ
fuelectenerg-13681	104	5	comparative	comparative	ADJ
fuelectenerg-13681	104	6	approaches	approach	NOUN
fuelectenerg-13681	104	7	a	a	DET
fuelectenerg-13681	104	8	limited	limited	ADJ
fuelectenerg-13681	104	9	number	number	NOUN
fuelectenerg-13681	104	10	of	of	ADP
fuelectenerg-13681	104	11	studies	study	NOUN
fuelectenerg-13681	104	12	have	have	AUX
fuelectenerg-13681	104	13	attempted	attempt	VERB
fuelectenerg-13681	104	14	to	to	PART
fuelectenerg-13681	104	15	bridge	bridge	VERB
fuelectenerg-13681	104	16	traditional	traditional	ADJ
fuelectenerg-13681	104	17	and	and	CCONJ
fuelectenerg-13681	104	18	learning	learning	NOUN
fuelectenerg-13681	104	19	-	-	PUNCT
fuelectenerg-13681	104	20	based	base	VERB
fuelectenerg-13681	104	21	methods	method	NOUN
fuelectenerg-13681	104	22	.	.	PUNCT
fuelectenerg-13681	105	1	xu	xu	PROPN
fuelectenerg-13681	106	1	and	and	CCONJ
fuelectenerg-13681	106	2	gao	gao	PROPN
fuelectenerg-13681	106	3	[	[	X
fuelectenerg-13681	106	4	30	30	NUM
fuelectenerg-13681	106	5	]	]	PUNCT
fuelectenerg-13681	106	6	proposed	propose	VERB
fuelectenerg-13681	106	7	a	a	DET
fuelectenerg-13681	106	8	model	model	NOUN
fuelectenerg-13681	106	9	aided	aid	VERB
fuelectenerg-13681	106	10	deep	deep	ADJ
fuelectenerg-13681	106	11	learning	learning	NOUN
fuelectenerg-13681	106	12	based	base	VERB
fuelectenerg-13681	106	13	mimo	mimo	PROPN
fuelectenerg-13681	106	14	ofdm	ofdm	PROPN
fuelectenerg-13681	106	15	receiver	receiver	NOUN
fuelectenerg-13681	106	16	that	that	PRON
fuelectenerg-13681	106	17	combined	combine	VERB
fuelectenerg-13681	106	18	traditional	traditional	ADJ
fuelectenerg-13681	106	19	least	least	ADJ
fuelectenerg-13681	106	20	square	square	ADJ
fuelectenerg-13681	106	21	channel	channel	NOUN
fuelectenerg-13681	106	22	estimation	estimation	NOUN
fuelectenerg-13681	106	23	with	with	ADP
fuelectenerg-13681	106	24	deep	deep	ADJ
fuelectenerg-13681	106	25	learning	learning	NOUN
fuelectenerg-13681	106	26	to	to	PART
fuelectenerg-13681	106	27	handle	handle	VERB
fuelectenerg-13681	106	28	nonlinear	nonlinear	ADJ
fuelectenerg-13681	106	29	power	power	NOUN
fuelectenerg-13681	106	30	amplifier	amplifier	NOUN
fuelectenerg-13681	106	31	distortions	distortion	NOUN
fuelectenerg-13681	106	32	.	.	PUNCT
fuelectenerg-13681	107	1	their	their	PRON
fuelectenerg-13681	107	2	approach	approach	NOUN
fuelectenerg-13681	107	3	demonstrated	demonstrate	VERB
fuelectenerg-13681	107	4	superior	superior	ADJ
fuelectenerg-13681	107	5	performance	performance	NOUN
fuelectenerg-13681	107	6	in	in	ADP
fuelectenerg-13681	107	7	terms	term	NOUN
fuelectenerg-13681	107	8	of	of	ADP
fuelectenerg-13681	107	9	bit	bit	NOUN
fuelectenerg-13681	107	10	error	error	NOUN
fuelectenerg-13681	107	11	rate	rate	NOUN
fuelectenerg-13681	107	12	and	and	CCONJ
fuelectenerg-13681	107	13	robustness	robustness	NOUN
fuelectenerg-13681	107	14	across	across	ADP
fuelectenerg-13681	107	15	varying	vary	VERB
fuelectenerg-13681	107	16	distortion	distortion	NOUN
fuelectenerg-13681	107	17	levels	level	NOUN
fuelectenerg-13681	107	18	,	,	PUNCT
fuelectenerg-13681	107	19	showing	show	VERB
fuelectenerg-13681	107	20	the	the	DET
fuelectenerg-13681	107	21	complementary	complementary	ADJ
fuelectenerg-13681	107	22	benefits	benefit	NOUN
fuelectenerg-13681	107	23	of	of	ADP
fuelectenerg-13681	107	24	combining	combine	VERB
fuelectenerg-13681	107	25	model	model	NOUN
fuelectenerg-13681	107	26	-	-	PUNCT
fuelectenerg-13681	107	27	based	base	VERB
fuelectenerg-13681	107	28	and	and	CCONJ
fuelectenerg-13681	107	29	data	data	NOUN
fuelectenerg-13681	107	30	-	-	PUNCT
fuelectenerg-13681	107	31	driven	drive	VERB
fuelectenerg-13681	107	32	techniques	technique	NOUN
fuelectenerg-13681	107	33	.	.	PUNCT
fuelectenerg-13681	108	1	comparative	comparative	ADJ
fuelectenerg-13681	108	2	analyses	analysis	NOUN
fuelectenerg-13681	108	3	between	between	ADP
fuelectenerg-13681	108	4	traditional	traditional	ADJ
fuelectenerg-13681	108	5	and	and	CCONJ
fuelectenerg-13681	108	6	learning	learning	NOUN
fuelectenerg-13681	108	7	-	-	PUNCT
fuelectenerg-13681	108	8	based	base	VERB
fuelectenerg-13681	108	9	approaches	approach	NOUN
fuelectenerg-13681	108	10	remain	remain	VERB
fuelectenerg-13681	108	11	scarce	scarce	ADJ
fuelectenerg-13681	108	12	.	.	PUNCT
fuelectenerg-13681	109	1	wu	wu	PROPN
fuelectenerg-13681	109	2	et	et	PROPN
fuelectenerg-13681	109	3	al	al	PROPN
fuelectenerg-13681	109	4	.	.	PUNCT
fuelectenerg-13681	110	1	[	[	X
fuelectenerg-13681	110	2	29	29	NUM
fuelectenerg-13681	110	3	]	]	PUNCT
fuelectenerg-13681	110	4	conducted	conduct	VERB
fuelectenerg-13681	110	5	a	a	DET
fuelectenerg-13681	110	6	limited	limited	ADJ
fuelectenerg-13681	110	7	comparison	comparison	NOUN
fuelectenerg-13681	110	8	between	between	ADP
fuelectenerg-13681	110	9	dnn	dnn	PROPN
fuelectenerg-13681	110	10	-	-	PUNCT
fuelectenerg-13681	110	11	based	base	VERB
fuelectenerg-13681	110	12	and	and	CCONJ
fuelectenerg-13681	110	13	mmse	mmse	ADJ
fuelectenerg-13681	110	14	beamforming	beamforme	VERB
fuelectenerg-13681	110	15	in	in	ADP
fuelectenerg-13681	110	16	simulated	simulated	ADJ
fuelectenerg-13681	110	17	environments	environment	NOUN
fuelectenerg-13681	110	18	.	.	PUNCT
fuelectenerg-13681	111	1	their	their	PRON
fuelectenerg-13681	111	2	results	result	NOUN
fuelectenerg-13681	111	3	indicated	indicate	VERB
fuelectenerg-13681	111	4	that	that	SCONJ
fuelectenerg-13681	111	5	deep	deep	ADJ
fuelectenerg-13681	111	6	learning	learning	NOUN
fuelectenerg-13681	111	7	approaches	approach	NOUN
fuelectenerg-13681	111	8	were	be	AUX
fuelectenerg-13681	111	9	more	more	ADV
fuelectenerg-13681	111	10	robust	robust	ADJ
fuelectenerg-13681	111	11	to	to	PART
fuelectenerg-13681	111	12	channel	channel	VERB
fuelectenerg-13681	111	13	estimation	estimation	NOUN
fuelectenerg-13681	111	14	errors	error	NOUN
fuelectenerg-13681	111	15	but	but	CCONJ
fuelectenerg-13681	111	16	provided	provide	VERB
fuelectenerg-13681	111	17	limited	limited	ADJ
fuelectenerg-13681	111	18	insights	insight	NOUN
fuelectenerg-13681	111	19	into	into	ADP
fuelectenerg-13681	111	20	computational	computational	ADJ
fuelectenerg-13681	111	21	aspects	aspect	NOUN
fuelectenerg-13681	111	22	and	and	CCONJ
fuelectenerg-13681	111	23	adaptation	adaptation	NOUN
fuelectenerg-13681	111	24	capabilities	capability	NOUN
fuelectenerg-13681	111	25	.	.	PUNCT
fuelectenerg-13681	112	1	for	for	ADP
fuelectenerg-13681	112	2	deep	deep	ADJ
fuelectenerg-13681	112	3	learning	learning	NOUN
fuelectenerg-13681	112	4	approaches	approach	NOUN
fuelectenerg-13681	112	5	,	,	PUNCT
fuelectenerg-13681	112	6	theoretical	theoretical	ADJ
fuelectenerg-13681	112	7	performance	performance	NOUN
fuelectenerg-13681	112	8	analysis	analysis	NOUN
fuelectenerg-13681	112	9	has	have	AUX
fuelectenerg-13681	112	10	been	be	AUX
fuelectenerg-13681	112	11	more	more	ADV
fuelectenerg-13681	112	12	challenging	challenging	ADJ
fuelectenerg-13681	112	13	.	.	PUNCT
fuelectenerg-13681	113	1	zappone	zappone	NOUN
fuelectenerg-13681	113	2	et	et	NOUN
fuelectenerg-13681	113	3	al	al	PROPN
fuelectenerg-13681	113	4	.	.	PUNCT
fuelectenerg-13681	114	1	[	[	X
fuelectenerg-13681	114	2	32	32	NUM
fuelectenerg-13681	114	3	]	]	PUNCT
fuelectenerg-13681	114	4	recently	recently	ADV
fuelectenerg-13681	114	5	developed	develop	VERB
fuelectenerg-13681	114	6	a	a	DET
fuelectenerg-13681	114	7	framework	framework	NOUN
fuelectenerg-13681	114	8	for	for	ADP
fuelectenerg-13681	114	9	analyzing	analyze	VERB
fuelectenerg-13681	114	10	wireless	wireless	ADJ
fuelectenerg-13681	114	11	network	network	NOUN
fuelectenerg-13681	114	12	design	design	NOUN
fuelectenerg-13681	114	13	in	in	ADP
fuelectenerg-13681	114	14	the	the	DET
fuelectenerg-13681	114	15	era	era	NOUN
fuelectenerg-13681	114	16	of	of	ADP
fuelectenerg-13681	114	17	deep	deep	ADJ
fuelectenerg-13681	114	18	learning	learning	NOUN
fuelectenerg-13681	114	19	,	,	PUNCT
fuelectenerg-13681	114	20	comparing	compare	VERB
fuelectenerg-13681	114	21	model	model	NOUN
fuelectenerg-13681	114	22	-	-	PUNCT
fuelectenerg-13681	114	23	based	base	VERB
fuelectenerg-13681	114	24	and	and	CCONJ
fuelectenerg-13681	114	25	ai	ai	NOUN
fuelectenerg-13681	114	26	-	-	PUNCT
fuelectenerg-13681	114	27	based	base	VERB
fuelectenerg-13681	114	28	approaches	approach	NOUN
fuelectenerg-13681	114	29	.	.	PUNCT
fuelectenerg-13681	115	1	they	they	PRON
fuelectenerg-13681	115	2	established	establish	VERB
fuelectenerg-13681	115	3	bounds	bound	NOUN
fuelectenerg-13681	115	4	on	on	ADP
fuelectenerg-13681	115	5	the	the	DET
fuelectenerg-13681	115	6	expected	expect	VERB
fuelectenerg-13681	115	7	error	error	NOUN
fuelectenerg-13681	115	8	as	as	ADP
fuelectenerg-13681	115	9	a	a	DET
fuelectenerg-13681	115	10	function	function	NOUN
fuelectenerg-13681	115	11	of	of	ADP
fuelectenerg-13681	115	12	training	training	NOUN
fuelectenerg-13681	115	13	dataset	dataset	NOUN
fuelectenerg-13681	115	14	size	size	NOUN
fuelectenerg-13681	115	15	,	,	PUNCT
fuelectenerg-13681	115	16	model	model	NOUN
fuelectenerg-13681	115	17	complexity	complexity	NOUN
fuelectenerg-13681	115	18	,	,	PUNCT
fuelectenerg-13681	115	19	and	and	CCONJ
fuelectenerg-13681	115	20	channel	channel	NOUN
fuelectenerg-13681	115	21	characteristics	characteristic	NOUN
fuelectenerg-13681	115	22	.	.	PUNCT
fuelectenerg-13681	116	1	the	the	DET
fuelectenerg-13681	116	2	convergence	convergence	NOUN
fuelectenerg-13681	116	3	properties	property	NOUN
fuelectenerg-13681	116	4	of	of	ADP
fuelectenerg-13681	116	5	traditional	traditional	ADJ
fuelectenerg-13681	116	6	optimization	optimization	NOUN
fuelectenerg-13681	116	7	algorithms	algorithm	NOUN
fuelectenerg-13681	116	8	are	be	AUX
fuelectenerg-13681	116	9	well	well	ADV
fuelectenerg-13681	116	10	-	-	PUNCT
fuelectenerg-13681	116	11	understood	understand	VERB
fuelectenerg-13681	116	12	.	.	PUNCT
fuelectenerg-13681	117	1	song	song	NOUN
fuelectenerg-13681	117	2	et	et	PROPN
fuelectenerg-13681	117	3	al	al	PROPN
fuelectenerg-13681	117	4	.	.	PUNCT
fuelectenerg-13681	118	1	[	[	X
fuelectenerg-13681	118	2	33	33	NUM
fuelectenerg-13681	118	3	]	]	PUNCT
fuelectenerg-13681	118	4	applied	apply	VERB
fuelectenerg-13681	118	5	adaptive	adaptive	ADJ
fuelectenerg-13681	118	6	filter	filter	NOUN
fuelectenerg-13681	118	7	theory	theory	NOUN
fuelectenerg-13681	118	8	to	to	PART
fuelectenerg-13681	118	9	harmonic	harmonic	VERB
fuelectenerg-13681	118	10	current	current	ADJ
fuelectenerg-13681	118	11	202	202	NUM
fuelectenerg-13681	118	12	m.	m.	NOUN
fuelectenerg-13681	118	13	mohamed	mohamed	PROPN
fuelectenerg-13681	118	14	,	,	PUNCT
fuelectenerg-13681	118	15	f.	f.	PROPN
fuelectenerg-13681	118	16	aljuaid	aljuaid	PROPN
fuelectenerg-13681	118	17	,	,	PUNCT
fuelectenerg-13681	118	18	mhd	mhd	PROPN
fuelectenerg-13681	118	19	w.	w.	PROPN
fuelectenerg-13681	118	20	koubeisi	koubeisi	PROPN
fuelectenerg-13681	118	21	suppression	suppression	NOUN
fuelectenerg-13681	118	22	,	,	PUNCT
fuelectenerg-13681	118	23	demonstrating	demonstrate	VERB
fuelectenerg-13681	118	24	principles	principle	NOUN
fuelectenerg-13681	118	25	applicable	applicable	ADJ
fuelectenerg-13681	118	26	to	to	ADP
fuelectenerg-13681	118	27	beamforming	beamforme	VERB
fuelectenerg-13681	118	28	.	.	PUNCT
fuelectenerg-13681	119	1	they	they	PRON
fuelectenerg-13681	119	2	established	establish	VERB
fuelectenerg-13681	119	3	that	that	SCONJ
fuelectenerg-13681	119	4	with	with	ADP
fuelectenerg-13681	119	5	properly	properly	ADV
fuelectenerg-13681	119	6	selected	select	VERB
fuelectenerg-13681	119	7	parameters	parameter	NOUN
fuelectenerg-13681	119	8	,	,	PUNCT
fuelectenerg-13681	119	9	these	these	DET
fuelectenerg-13681	119	10	methods	method	NOUN
fuelectenerg-13681	119	11	can	can	AUX
fuelectenerg-13681	119	12	achieve	achieve	VERB
fuelectenerg-13681	119	13	optimal	optimal	ADJ
fuelectenerg-13681	119	14	performance	performance	NOUN
fuelectenerg-13681	119	15	.	.	PUNCT
fuelectenerg-13681	120	1	for	for	ADP
fuelectenerg-13681	120	2	deep	deep	ADJ
fuelectenerg-13681	120	3	learning	learning	NOUN
fuelectenerg-13681	120	4	approaches	approach	NOUN
fuelectenerg-13681	120	5	,	,	PUNCT
fuelectenerg-13681	120	6	convergence	convergence	NOUN
fuelectenerg-13681	120	7	analysis	analysis	NOUN
fuelectenerg-13681	120	8	typically	typically	ADV
fuelectenerg-13681	120	9	focuses	focus	VERB
fuelectenerg-13681	120	10	on	on	ADP
fuelectenerg-13681	120	11	the	the	DET
fuelectenerg-13681	120	12	training	training	NOUN
fuelectenerg-13681	120	13	process	process	NOUN
fuelectenerg-13681	120	14	.	.	PUNCT
fuelectenerg-13681	121	1	rodriguez	rodriguez	NOUN
fuelectenerg-13681	121	2	et	et	PROPN
fuelectenerg-13681	121	3	al	al	PROPN
fuelectenerg-13681	121	4	.	.	PUNCT
fuelectenerg-13681	122	1	[	[	X
fuelectenerg-13681	122	2	34	34	NUM
fuelectenerg-13681	122	3	]	]	PUNCT
fuelectenerg-13681	122	4	analyzed	analyze	VERB
fuelectenerg-13681	122	5	deep	deep	ADJ
fuelectenerg-13681	122	6	learning	learning	NOUN
fuelectenerg-13681	122	7	techniques	technique	NOUN
fuelectenerg-13681	122	8	for	for	ADP
fuelectenerg-13681	122	9	cybersecurity	cybersecurity	NOUN
fuelectenerg-13681	122	10	in	in	ADP
fuelectenerg-13681	122	11	mobile	mobile	ADJ
fuelectenerg-13681	122	12	networks	network	NOUN
fuelectenerg-13681	122	13	,	,	PUNCT
fuelectenerg-13681	122	14	presenting	present	VERB
fuelectenerg-13681	122	15	insights	insight	NOUN
fuelectenerg-13681	122	16	also	also	ADV
fuelectenerg-13681	122	17	relevant	relevant	ADJ
fuelectenerg-13681	122	18	to	to	PART
fuelectenerg-13681	122	19	signal	signal	VERB
fuelectenerg-13681	122	20	processing	processing	NOUN
fuelectenerg-13681	122	21	tasks	task	NOUN
fuelectenerg-13681	122	22	.	.	PUNCT
fuelectenerg-13681	123	1	they	they	PRON
fuelectenerg-13681	123	2	established	establish	VERB
fuelectenerg-13681	123	3	that	that	SCONJ
fuelectenerg-13681	123	4	with	with	ADP
fuelectenerg-13681	123	5	properly	properly	ADV
fuelectenerg-13681	123	6	selected	select	VERB
fuelectenerg-13681	123	7	learning	learning	NOUN
fuelectenerg-13681	123	8	rates	rate	NOUN
fuelectenerg-13681	123	9	,	,	PUNCT
fuelectenerg-13681	123	10	these	these	DET
fuelectenerg-13681	123	11	networks	network	NOUN
fuelectenerg-13681	123	12	can	can	AUX
fuelectenerg-13681	123	13	achieve	achieve	VERB
fuelectenerg-13681	123	14	an	an	DET
fuelectenerg-13681	123	15	εoptimal	εoptimal	ADJ
fuelectenerg-13681	123	16	solution	solution	NOUN
fuelectenerg-13681	123	17	with	with	ADP
fuelectenerg-13681	123	18	high	high	ADJ
fuelectenerg-13681	123	19	probability	probability	NOUN
fuelectenerg-13681	123	20	after	after	ADP
fuelectenerg-13681	123	21	sufficient	sufficient	ADJ
fuelectenerg-13681	123	22	iterations	iteration	NOUN
fuelectenerg-13681	123	23	.	.	PUNCT
fuelectenerg-13681	124	1	these	these	DET
fuelectenerg-13681	124	2	theoretical	theoretical	ADJ
fuelectenerg-13681	124	3	foundations	foundation	NOUN
fuelectenerg-13681	124	4	provide	provide	VERB
fuelectenerg-13681	124	5	essential	essential	ADJ
fuelectenerg-13681	124	6	context	context	NOUN
fuelectenerg-13681	124	7	for	for	ADP
fuelectenerg-13681	124	8	our	our	PRON
fuelectenerg-13681	124	9	comparative	comparative	ADJ
fuelectenerg-13681	124	10	analysis	analysis	NOUN
fuelectenerg-13681	124	11	,	,	PUNCT
fuelectenerg-13681	124	12	enabling	enable	VERB
fuelectenerg-13681	124	13	us	we	PRON
fuelectenerg-13681	124	14	to	to	PART
fuelectenerg-13681	124	15	not	not	PART
fuelectenerg-13681	124	16	only	only	ADV
fuelectenerg-13681	124	17	empirically	empirically	ADV
fuelectenerg-13681	124	18	evaluate	evaluate	VERB
fuelectenerg-13681	124	19	performance	performance	NOUN
fuelectenerg-13681	124	20	but	but	CCONJ
fuelectenerg-13681	124	21	also	also	ADV
fuelectenerg-13681	124	22	understand	understand	VERB
fuelectenerg-13681	124	23	the	the	DET
fuelectenerg-13681	124	24	fundamental	fundamental	ADJ
fuelectenerg-13681	124	25	limitations	limitation	NOUN
fuelectenerg-13681	124	26	of	of	ADP
fuelectenerg-13681	124	27	each	each	DET
fuelectenerg-13681	124	28	approach	approach	NOUN
fuelectenerg-13681	124	29	.	.	PUNCT
fuelectenerg-13681	125	1	3	3	X
fuelectenerg-13681	125	2	.	.	X
fuelectenerg-13681	125	3	methodology	methodology	NOUN
fuelectenerg-13681	125	4	this	this	DET
fuelectenerg-13681	125	5	section	section	NOUN
fuelectenerg-13681	125	6	presents	present	VERB
fuelectenerg-13681	125	7	our	our	PRON
fuelectenerg-13681	125	8	methodology	methodology	NOUN
fuelectenerg-13681	125	9	for	for	ADP
fuelectenerg-13681	125	10	comparatively	comparatively	ADV
fuelectenerg-13681	125	11	analyzing	analyze	VERB
fuelectenerg-13681	125	12	deep	deep	ADJ
fuelectenerg-13681	125	13	learning	learning	NOUN
fuelectenerg-13681	125	14	and	and	CCONJ
fuelectenerg-13681	125	15	traditional	traditional	ADJ
fuelectenerg-13681	125	16	optimization	optimization	NOUN
fuelectenerg-13681	125	17	algorithms	algorithm	NOUN
fuelectenerg-13681	125	18	for	for	ADP
fuelectenerg-13681	125	19	adaptive	adaptive	ADJ
fuelectenerg-13681	125	20	beamforming	beamforming	NOUN
fuelectenerg-13681	125	21	in	in	ADP
fuelectenerg-13681	125	22	mimo	mimo	PROPN
fuelectenerg-13681	125	23	systems	system	NOUN
fuelectenerg-13681	125	24	.	.	PUNCT
fuelectenerg-13681	126	1	we	we	PRON
fuelectenerg-13681	126	2	describe	describe	VERB
fuelectenerg-13681	126	3	the	the	DET
fuelectenerg-13681	126	4	system	system	NOUN
fuelectenerg-13681	126	5	model	model	NOUN
fuelectenerg-13681	126	6	,	,	PUNCT
fuelectenerg-13681	126	7	implementation	implementation	NOUN
fuelectenerg-13681	126	8	details	detail	NOUN
fuelectenerg-13681	126	9	of	of	ADP
fuelectenerg-13681	126	10	evaluated	evaluated	ADJ
fuelectenerg-13681	126	11	algorithms	algorithm	NOUN
fuelectenerg-13681	126	12	,	,	PUNCT
fuelectenerg-13681	126	13	dataset	dataset	ADJ
fuelectenerg-13681	126	14	preparation	preparation	NOUN
fuelectenerg-13681	126	15	,	,	PUNCT
fuelectenerg-13681	126	16	evaluation	evaluation	NOUN
fuelectenerg-13681	126	17	metrics	metric	NOUN
fuelectenerg-13681	126	18	,	,	PUNCT
fuelectenerg-13681	126	19	and	and	CCONJ
fuelectenerg-13681	126	20	experimental	experimental	ADJ
fuelectenerg-13681	126	21	setup	setup	NOUN
fuelectenerg-13681	126	22	.	.	PUNCT
fuelectenerg-13681	127	1	3.1	3.1	NUM
fuelectenerg-13681	127	2	.	.	PUNCT
fuelectenerg-13681	127	3	system	system	NOUN
fuelectenerg-13681	127	4	model	model	NOUN
fuelectenerg-13681	127	5	we	we	PRON
fuelectenerg-13681	127	6	consider	consider	VERB
fuelectenerg-13681	127	7	a	a	DET
fuelectenerg-13681	127	8	mimo	mimo	PROPN
fuelectenerg-13681	127	9	communication	communication	NOUN
fuelectenerg-13681	127	10	system	system	NOUN
fuelectenerg-13681	127	11	with	with	ADP
fuelectenerg-13681	127	12	nt	not	PART
fuelectenerg-13681	127	13	transmit	transmit	VERB
fuelectenerg-13681	127	14	antennas	antenna	NOUN
fuelectenerg-13681	127	15	and	and	CCONJ
fuelectenerg-13681	127	16	nr	nr	PROPN
fuelectenerg-13681	127	17	receive	receive	VERB
fuelectenerg-13681	127	18	antennas	antenna	NOUN
fuelectenerg-13681	127	19	.	.	PUNCT
fuelectenerg-13681	128	1	the	the	DET
fuelectenerg-13681	128	2	received	receive	VERB
fuelectenerg-13681	128	3	signal	signal	NOUN
fuelectenerg-13681	128	4	vector	vector	NOUN
fuelectenerg-13681	128	5	1rn	1rn	NOUN
fuelectenerg-13681	128	6			NOUN
fuelectenerg-13681	128	7	y	y	PRON
fuelectenerg-13681	128	8	is	be	AUX
fuelectenerg-13681	128	9	modeled	model	VERB
fuelectenerg-13681	128	10	as	as	ADP
fuelectenerg-13681	128	11	:	:	PUNCT
fuelectenerg-13681	128	12	=	=	SYM
fuelectenerg-13681	129	1	+	+	NOUN
fuelectenerg-13681	129	2	y	y	PROPN
fuelectenerg-13681	129	3	hwx	hwx	VERB
fuelectenerg-13681	129	4	n	n	CCONJ
fuelectenerg-13681	129	5	(	(	PUNCT
fuelectenerg-13681	129	6	1	1	X
fuelectenerg-13681	129	7	)	)	PUNCT
fuelectenerg-13681	129	8	where	where	SCONJ
fuelectenerg-13681	129	9	r	r	NOUN
fuelectenerg-13681	129	10	tn	tn	PROPN
fuelectenerg-13681	129	11	n	n	VERB
fuelectenerg-13681	129	12	h	h	NUM
fuelectenerg-13681	129	13	is	be	AUX
fuelectenerg-13681	129	14	the	the	DET
fuelectenerg-13681	129	15	channel	channel	NOUN
fuelectenerg-13681	129	16	matrix	matrix	NOUN
fuelectenerg-13681	129	17	,	,	PUNCT
fuelectenerg-13681	129	18	t	t	PROPN
fuelectenerg-13681	129	19	sn	sn	PROPN
fuelectenerg-13681	129	20	n	n	ADJ
fuelectenerg-13681	129	21	w	w	NOUN
fuelectenerg-13681	129	22	is	be	AUX
fuelectenerg-13681	129	23	the	the	DET
fuelectenerg-13681	129	24	beamforming	beamforme	VERB
fuelectenerg-13681	129	25	matrix	matrix	NOUN
fuelectenerg-13681	129	26	,	,	PUNCT
fuelectenerg-13681	129	27	1sn	1sn	PROPN
fuelectenerg-13681	129	28			PROPN
fuelectenerg-13681	129	29	x	x	PUNCT
fuelectenerg-13681	129	30	is	be	AUX
fuelectenerg-13681	129	31	the	the	DET
fuelectenerg-13681	129	32	transmitted	transmit	VERB
fuelectenerg-13681	129	33	signal	signal	NOUN
fuelectenerg-13681	129	34	vector	vector	NOUN
fuelectenerg-13681	129	35	with	with	ADP
fuelectenerg-13681	129	36	ns	ns	ADJ
fuelectenerg-13681	129	37	data	datum	NOUN
fuelectenerg-13681	129	38	streams	stream	NOUN
fuelectenerg-13681	129	39	,	,	PUNCT
fuelectenerg-13681	129	40	and	and	CCONJ
fuelectenerg-13681	129	41	1rn	1rn	ADJ
fuelectenerg-13681	129	42			PROPN
fuelectenerg-13681	129	43	n	n	PROPN
fuelectenerg-13681	129	44	is	be	AUX
fuelectenerg-13681	129	45	the	the	DET
fuelectenerg-13681	129	46	additive	additive	ADJ
fuelectenerg-13681	129	47	white	white	ADJ
fuelectenerg-13681	129	48	gaussian	gaussian	PROPN
fuelectenerg-13681	129	49	noise	noise	NOUN
fuelectenerg-13681	129	50	(	(	PUNCT
fuelectenerg-13681	129	51	awgn	awgn	NOUN
fuelectenerg-13681	129	52	)	)	PUNCT
fuelectenerg-13681	129	53	vector	vector	NOUN
fuelectenerg-13681	129	54	with	with	ADP
fuelectenerg-13681	129	55	covariance	covariance	NOUN
fuelectenerg-13681	129	56	matrix	matrix	NOUN
fuelectenerg-13681	129	57	2i	2i	NOUN
fuelectenerg-13681	129	58	.	.	PUNCT
fuelectenerg-13681	130	1	the	the	DET
fuelectenerg-13681	130	2	objective	objective	NOUN
fuelectenerg-13681	130	3	of	of	ADP
fuelectenerg-13681	130	4	adaptive	adaptive	ADJ
fuelectenerg-13681	130	5	beamforming	beamforming	NOUN
fuelectenerg-13681	130	6	is	be	AUX
fuelectenerg-13681	130	7	to	to	PART
fuelectenerg-13681	130	8	determine	determine	VERB
fuelectenerg-13681	130	9	the	the	DET
fuelectenerg-13681	130	10	optimal	optimal	ADJ
fuelectenerg-13681	130	11	beamforming	beamforming	NOUN
fuelectenerg-13681	130	12	matrix	matrix	NOUN
fuelectenerg-13681	130	13	w	w	NOUN
fuelectenerg-13681	130	14	that	that	PRON
fuelectenerg-13681	130	15	maximizes	maximize	VERB
fuelectenerg-13681	130	16	a	a	DET
fuelectenerg-13681	130	17	specific	specific	ADJ
fuelectenerg-13681	130	18	performance	performance	NOUN
fuelectenerg-13681	130	19	metric	metric	NOUN
fuelectenerg-13681	130	20	,	,	PUNCT
fuelectenerg-13681	130	21	typically	typically	ADV
fuelectenerg-13681	130	22	the	the	DET
fuelectenerg-13681	130	23	signal	signal	NOUN
fuelectenerg-13681	130	24	-	-	PUNCT
fuelectenerg-13681	130	25	to	to	ADP
fuelectenerg-13681	130	26	-	-	PUNCT
fuelectenerg-13681	130	27	interference	interference	NOUN
fuelectenerg-13681	130	28	-	-	PUNCT
fuelectenerg-13681	130	29	plus	plus	CCONJ
fuelectenerg-13681	130	30	-	-	PUNCT
fuelectenerg-13681	130	31	noise	noise	NOUN
fuelectenerg-13681	130	32	ratio	ratio	NOUN
fuelectenerg-13681	130	33	(	(	PUNCT
fuelectenerg-13681	130	34	sinr	sinr	NOUN
fuelectenerg-13681	130	35	)	)	PUNCT
fuelectenerg-13681	130	36	,	,	PUNCT
fuelectenerg-13681	130	37	under	under	ADP
fuelectenerg-13681	130	38	various	various	ADJ
fuelectenerg-13681	130	39	constraints	constraint	NOUN
fuelectenerg-13681	130	40	.	.	PUNCT
fuelectenerg-13681	131	1	for	for	ADP
fuelectenerg-13681	131	2	single	single	ADJ
fuelectenerg-13681	131	3	-	-	PUNCT
fuelectenerg-13681	131	4	user	user	NOUN
fuelectenerg-13681	131	5	mimo	mimo	PROPN
fuelectenerg-13681	131	6	systems	systems	PROPN
fuelectenerg-13681	131	7	,	,	PUNCT
fuelectenerg-13681	131	8	this	this	PRON
fuelectenerg-13681	131	9	can	can	AUX
fuelectenerg-13681	131	10	be	be	AUX
fuelectenerg-13681	131	11	formulated	formulate	VERB
fuelectenerg-13681	131	12	as	as	ADP
fuelectenerg-13681	131	13	:	:	SYM
fuelectenerg-13681	131	14	2	2	NUM
fuelectenerg-13681	131	15	2	2	NUM
fuelectenerg-13681	131	16	max2	max2	NOUN
fuelectenerg-13681	131	17	|	|	ADV
fuelectenerg-13681	131	18	|	|	ADV
fuelectenerg-13681	131	19	maxsinr	maxsinr	VERB
fuelectenerg-13681	131	20	    	    	SPACE
fuelectenerg-13681	131	21	s.t	s.t	PROPN
fuelectenerg-13681	131	22	.	.	PROPN
fuelectenerg-13681	131	23	    	    	SPACE
fuelectenerg-13681	132	1	|	|	ADV
fuelectenerg-13681	132	2	|	|	ADV
fuelectenerg-13681	132	3	|	|	ADV
fuelectenerg-13681	132	4	|	|	ADV
fuelectenerg-13681	132	5	p=	p=	VERB
fuelectenerg-13681	132	6			PROPN
fuelectenerg-13681	132	7	w	w	NOUN
fuelectenerg-13681	132	8	hwx	hwx	NOUN
fuelectenerg-13681	132	9	w	w	PROPN
fuelectenerg-13681	132	10	n	n	PROPN
fuelectenerg-13681	132	11	(	(	PUNCT
fuelectenerg-13681	132	12	2	2	NUM
fuelectenerg-13681	132	13	)	)	PUNCT
fuelectenerg-13681	132	14	where	where	SCONJ
fuelectenerg-13681	132	15	pmax	pmax	NOUN
fuelectenerg-13681	132	16	is	be	AUX
fuelectenerg-13681	132	17	the	the	DET
fuelectenerg-13681	132	18	maximum	maximum	ADJ
fuelectenerg-13681	132	19	transmit	transmit	NOUN
fuelectenerg-13681	132	20	power	power	NOUN
fuelectenerg-13681	132	21	constraint	constraint	NOUN
fuelectenerg-13681	132	22	.	.	PUNCT
fuelectenerg-13681	133	1	for	for	ADP
fuelectenerg-13681	133	2	multi	multi	ADJ
fuelectenerg-13681	133	3	-	-	PROPN
fuelectenerg-13681	133	4	user	user	NOUN
fuelectenerg-13681	133	5	mimo	mimo	PROPN
fuelectenerg-13681	133	6	systems	systems	PROPN
fuelectenerg-13681	133	7	,	,	PUNCT
fuelectenerg-13681	133	8	additional	additional	ADJ
fuelectenerg-13681	133	9	constraints	constraint	NOUN
fuelectenerg-13681	133	10	are	be	AUX
fuelectenerg-13681	133	11	introduced	introduce	VERB
fuelectenerg-13681	133	12	to	to	PART
fuelectenerg-13681	133	13	minimize	minimize	VERB
fuelectenerg-13681	133	14	inter	inter	ADJ
fuelectenerg-13681	133	15	-	-	ADJ
fuelectenerg-13681	133	16	user	user	NOUN
fuelectenerg-13681	133	17	interference	interference	NOUN
fuelectenerg-13681	133	18	.	.	PUNCT
fuelectenerg-13681	134	1	the	the	DET
fuelectenerg-13681	134	2	optimal	optimal	ADJ
fuelectenerg-13681	134	3	beamforming	beamforming	NOUN
fuelectenerg-13681	134	4	matrix	matrix	NOUN
fuelectenerg-13681	134	5	w	w	NOUN
fuelectenerg-13681	134	6	*	*	NOUN
fuelectenerg-13681	134	7	that	that	PRON
fuelectenerg-13681	134	8	maximizes	maximize	VERB
fuelectenerg-13681	134	9	the	the	DET
fuelectenerg-13681	134	10	sinr	sinr	NOUN
fuelectenerg-13681	134	11	can	can	AUX
fuelectenerg-13681	134	12	be	be	AUX
fuelectenerg-13681	134	13	obtained	obtain	VERB
fuelectenerg-13681	134	14	through	through	ADP
fuelectenerg-13681	134	15	eigenvalue	eigenvalue	ADJ
fuelectenerg-13681	134	16	decomposition	decomposition	NOUN
fuelectenerg-13681	134	17	:	:	PUNCT
fuelectenerg-13681	135	1	*	*	PUNCT
fuelectenerg-13681	135	2	max	max	PROPN
fuelectenerg-13681	135	3	1p=	1p=	NUM
fuelectenerg-13681	135	4	w	w	VERB
fuelectenerg-13681	135	5	v	v	NOUN
fuelectenerg-13681	135	6	(	(	PUNCT
fuelectenerg-13681	135	7	3	3	NUM
fuelectenerg-13681	135	8	)	)	PUNCT
fuelectenerg-13681	135	9	where	where	SCONJ
fuelectenerg-13681	135	10	v1	v1	NOUN
fuelectenerg-13681	135	11	contains	contain	VERB
fuelectenerg-13681	135	12	the	the	DET
fuelectenerg-13681	135	13	eigenvectors	eigenvector	NOUN
fuelectenerg-13681	135	14	corresponding	correspond	VERB
fuelectenerg-13681	135	15	to	to	ADP
fuelectenerg-13681	135	16	the	the	DET
fuelectenerg-13681	135	17	largest	large	ADJ
fuelectenerg-13681	135	18	eigenvalues	eigenvalue	NOUN
fuelectenerg-13681	135	19	of	of	ADP
fuelectenerg-13681	135	20	hhh	hhh	PROPN
fuelectenerg-13681	135	21	.	.	PUNCT
fuelectenerg-13681	136	1	for	for	ADP
fuelectenerg-13681	136	2	time	time	NOUN
fuelectenerg-13681	136	3	-	-	PUNCT
fuelectenerg-13681	136	4	varying	vary	VERB
fuelectenerg-13681	136	5	channels	channel	NOUN
fuelectenerg-13681	136	6	,	,	PUNCT
fuelectenerg-13681	136	7	we	we	PRON
fuelectenerg-13681	136	8	model	model	VERB
fuelectenerg-13681	136	9	the	the	DET
fuelectenerg-13681	136	10	channel	channel	NOUN
fuelectenerg-13681	136	11	matrix	matrix	NOUN
fuelectenerg-13681	136	12	at	at	ADP
fuelectenerg-13681	136	13	time	time	NOUN
fuelectenerg-13681	136	14	t	t	PROPN
fuelectenerg-13681	136	15	is	be	AUX
fuelectenerg-13681	136	16	:	:	PUNCT
fuelectenerg-13681	136	17	2	2	NUM
fuelectenerg-13681	136	18	1	1	NUM
fuelectenerg-13681	136	19	new1	new1	PROPN
fuelectenerg-13681	136	20	t	t	PROPN
fuelectenerg-13681	136	21	t	t	PROPN
fuelectenerg-13681	136	22	−=	−=	X
fuelectenerg-13681	136	23	+	+	CCONJ
fuelectenerg-13681	136	24	−h	−h	ADJ
fuelectenerg-13681	136	25	h	h	NOUN
fuelectenerg-13681	136	26	h	h	NOUN
fuelectenerg-13681	136	27	(	(	PUNCT
fuelectenerg-13681	136	28	4	4	X
fuelectenerg-13681	136	29	)	)	PUNCT
fuelectenerg-13681	136	30	deep	deep	ADJ
fuelectenerg-13681	136	31	learning	learning	NOUN
fuelectenerg-13681	136	32	and	and	CCONJ
fuelectenerg-13681	136	33	traditional	traditional	ADJ
fuelectenerg-13681	136	34	algorithms	algorithm	NOUN
fuelectenerg-13681	136	35	for	for	ADP
fuelectenerg-13681	136	36	mimo	mimo	PROPN
fuelectenerg-13681	136	37	beamforming	beamforme	VERB
fuelectenerg-13681	136	38	203	203	NUM
fuelectenerg-13681	136	39	where	where	SCONJ
fuelectenerg-13681	136	40			VERB
fuelectenerg-13681	136	41			NOUN
fuelectenerg-13681	136	42	[	[	X
fuelectenerg-13681	136	43	0,1	0,1	NUM
fuelectenerg-13681	136	44	]	]	PUNCT
fuelectenerg-13681	136	45	is	be	AUX
fuelectenerg-13681	136	46	the	the	DET
fuelectenerg-13681	136	47	temporal	temporal	ADJ
fuelectenerg-13681	136	48	correlation	correlation	NOUN
fuelectenerg-13681	136	49	coefficient	coefficient	NOUN
fuelectenerg-13681	136	50	and	and	CCONJ
fuelectenerg-13681	136	51	hnew	hnew	ADJ
fuelectenerg-13681	136	52	is	be	AUX
fuelectenerg-13681	136	53	an	an	DET
fuelectenerg-13681	136	54	independent	independent	ADJ
fuelectenerg-13681	136	55	channel	channel	NOUN
fuelectenerg-13681	136	56	realization	realization	NOUN
fuelectenerg-13681	136	57	.	.	PUNCT
fuelectenerg-13681	137	1	this	this	DET
fuelectenerg-13681	137	2	model	model	NOUN
fuelectenerg-13681	137	3	allows	allow	VERB
fuelectenerg-13681	137	4	us	we	PRON
fuelectenerg-13681	137	5	to	to	PART
fuelectenerg-13681	137	6	systematically	systematically	ADV
fuelectenerg-13681	137	7	analyze	analyze	VERB
fuelectenerg-13681	137	8	algorithm	algorithm	NOUN
fuelectenerg-13681	137	9	performance	performance	NOUN
fuelectenerg-13681	137	10	under	under	ADP
fuelectenerg-13681	137	11	different	different	ADJ
fuelectenerg-13681	137	12	mobility	mobility	NOUN
fuelectenerg-13681	137	13	scenarios	scenario	NOUN
fuelectenerg-13681	137	14	.	.	PUNCT
fuelectenerg-13681	138	1	3.2	3.2	NUM
fuelectenerg-13681	138	2	.	.	PUNCT
fuelectenerg-13681	138	3	traditional	traditional	ADJ
fuelectenerg-13681	138	4	optimization	optimization	NOUN
fuelectenerg-13681	138	5	algorithms	algorithm	NOUN
fuelectenerg-13681	138	6	we	we	PRON
fuelectenerg-13681	138	7	implement	implement	VERB
fuelectenerg-13681	138	8	and	and	CCONJ
fuelectenerg-13681	138	9	evaluate	evaluate	VERB
fuelectenerg-13681	138	10	four	four	NUM
fuelectenerg-13681	138	11	representative	representative	ADJ
fuelectenerg-13681	138	12	traditional	traditional	ADJ
fuelectenerg-13681	138	13	optimization	optimization	NOUN
fuelectenerg-13681	138	14	algorithms	algorithm	NOUN
fuelectenerg-13681	138	15	for	for	ADP
fuelectenerg-13681	138	16	adaptive	adaptive	ADJ
fuelectenerg-13681	138	17	beamforming	beamforming	NOUN
fuelectenerg-13681	138	18	:	:	PUNCT
fuelectenerg-13681	138	19	least	least	ADJ
fuelectenerg-13681	138	20	mean	mean	VERB
fuelectenerg-13681	138	21	squares	square	NOUN
fuelectenerg-13681	138	22	(	(	PUNCT
fuelectenerg-13681	138	23	lms	lm	NOUN
fuelectenerg-13681	138	24	):	):	PUNCT
fuelectenerg-13681	138	25	*	*	PUNCT
fuelectenerg-13681	138	26	(	(	PUNCT
fuelectenerg-13681	138	27	)	)	PUNCT
fuelectenerg-13681	138	28	(	(	PUNCT
fuelectenerg-13681	138	29	1	1	NUM
fuelectenerg-13681	138	30	)	)	PUNCT
fuelectenerg-13681	138	31	(	(	PUNCT
fuelectenerg-13681	138	32	)	)	PUNCT
fuelectenerg-13681	138	33	(	(	PUNCT
fuelectenerg-13681	138	34	)	)	PUNCT
fuelectenerg-13681	138	35	n	n	CCONJ
fuelectenerg-13681	138	36	n	n	CCONJ
fuelectenerg-13681	138	37	n	n	ADV
fuelectenerg-13681	138	38	ne+	ne+	NOUN
fuelectenerg-13681	138	39	=	=	ADJ
fuelectenerg-13681	139	1	+	+	PROPN
fuelectenerg-13681	139	2	w	w	PROPN
fuelectenerg-13681	139	3	w	w	PROPN
fuelectenerg-13681	139	4	x	x	X
fuelectenerg-13681	139	5	(	(	PUNCT
fuelectenerg-13681	139	6	5	5	NUM
fuelectenerg-13681	139	7	)	)	PUNCT
fuelectenerg-13681	139	8	where	where	SCONJ
fuelectenerg-13681	139	9			NOUN
fuelectenerg-13681	139	10	is	be	AUX
fuelectenerg-13681	139	11	the	the	DET
fuelectenerg-13681	139	12	step	step	NOUN
fuelectenerg-13681	139	13	size	size	NOUN
fuelectenerg-13681	139	14	,	,	PUNCT
fuelectenerg-13681	139	15	x(n	x(n	NOUN
fuelectenerg-13681	139	16	)	)	PUNCT
fuelectenerg-13681	139	17	is	be	AUX
fuelectenerg-13681	139	18	the	the	DET
fuelectenerg-13681	139	19	input	input	NOUN
fuelectenerg-13681	139	20	signal	signal	PROPN
fuelectenerg-13681	139	21	vector	vector	NOUN
fuelectenerg-13681	139	22	,	,	PUNCT
fuelectenerg-13681	139	23	and	and	CCONJ
fuelectenerg-13681	139	24	e(n	e(n	NOUN
fuelectenerg-13681	139	25	)	)	PUNCT
fuelectenerg-13681	139	26	is	be	AUX
fuelectenerg-13681	139	27	the	the	DET
fuelectenerg-13681	139	28	error	error	NOUN
fuelectenerg-13681	139	29	between	between	ADP
fuelectenerg-13681	139	30	the	the	DET
fuelectenerg-13681	139	31	desired	desire	VERB
fuelectenerg-13681	139	32	and	and	CCONJ
fuelectenerg-13681	139	33	actual	actual	ADJ
fuelectenerg-13681	139	34	output	output	NOUN
fuelectenerg-13681	139	35	.	.	PUNCT
fuelectenerg-13681	140	1	the	the	DET
fuelectenerg-13681	140	2	convergence	convergence	NOUN
fuelectenerg-13681	140	3	rate	rate	NOUN
fuelectenerg-13681	140	4	of	of	ADP
fuelectenerg-13681	140	5	lms	lm	NOUN
fuelectenerg-13681	140	6	is	be	AUX
fuelectenerg-13681	140	7	governed	govern	VERB
fuelectenerg-13681	140	8	by	by	ADP
fuelectenerg-13681	140	9	the	the	DET
fuelectenerg-13681	140	10	eigenvalue	eigenvalue	ADJ
fuelectenerg-13681	140	11	spread	spread	NOUN
fuelectenerg-13681	140	12	of	of	ADP
fuelectenerg-13681	140	13	the	the	DET
fuelectenerg-13681	140	14	correlation	correlation	NOUN
fuelectenerg-13681	140	15	matrix	matrix	NOUN
fuelectenerg-13681	140	16	[	[	PUNCT
fuelectenerg-13681	140	17	(	(	PUNCT
fuelectenerg-13681	140	18	)	)	PUNCT
fuelectenerg-13681	140	19	(	(	PUNCT
fuelectenerg-13681	140	20	)	)	PUNCT
fuelectenerg-13681	140	21	]	]	X
fuelectenerg-13681	140	22	he	he	PRON
fuelectenerg-13681	140	23	n	n	CCONJ
fuelectenerg-13681	140	24	n	n	CCONJ
fuelectenerg-13681	140	25	=	=	NOUN
fuelectenerg-13681	140	26	r	r	NOUN
fuelectenerg-13681	140	27	x	x	NOUN
fuelectenerg-13681	140	28	x	x	X
fuelectenerg-13681	140	29	.	.	PUNCT
fuelectenerg-13681	141	1	specifically	specifically	ADV
fuelectenerg-13681	141	2	,	,	PUNCT
fuelectenerg-13681	141	3	the	the	DET
fuelectenerg-13681	141	4	mean	mean	ADJ
fuelectenerg-13681	141	5	square	square	ADJ
fuelectenerg-13681	141	6	error	error	NOUN
fuelectenerg-13681	141	7	converges	converge	VERB
fuelectenerg-13681	141	8	approximately	approximately	ADV
fuelectenerg-13681	141	9	as	as	ADP
fuelectenerg-13681	141	10	:	:	PUNCT
fuelectenerg-13681	141	11			PROPN
fuelectenerg-13681	141	12			PROPN
fuelectenerg-13681	141	13	min22	min22	PROPN
fuelectenerg-13681	141	14	2	2	NUM
fuelectenerg-13681	141	15	(	(	PUNCT
fuelectenerg-13681	141	16	)	)	PUNCT
fuelectenerg-13681	142	1	|	|	ADV
fuelectenerg-13681	142	2	(	(	PUNCT
fuelectenerg-13681	142	3	0	0	NUM
fuelectenerg-13681	142	4	)	)	PUNCT
fuelectenerg-13681	143	1	|	|	ADV
fuelectenerg-13681	143	2	n	n	NOUN
fuelectenerg-13681	143	3	e	e	NOUN
fuelectenerg-13681	143	4	e	e	X
fuelectenerg-13681	143	5	n	n	X
fuelectenerg-13681	143	6	e	e	X
fuelectenerg-13681	143	7	e	e	X
fuelectenerg-13681	143	8	e	e	ADP
fuelectenerg-13681	143	9	−	−	PROPN
fuelectenerg-13681	143	10			NOUN
fuelectenerg-13681	143	11			ADJ
fuelectenerg-13681	143	12			NOUN
fuelectenerg-13681	143	13			PROPN
fuelectenerg-13681	143	14	(	(	PUNCT
fuelectenerg-13681	143	15	6	6	NUM
fuelectenerg-13681	143	16	)	)	PUNCT
fuelectenerg-13681	143	17	where	where	SCONJ
fuelectenerg-13681	143	18	min	min	NOUN
fuelectenerg-13681	143	19	is	be	AUX
fuelectenerg-13681	143	20	the	the	DET
fuelectenerg-13681	143	21	minimum	minimum	ADJ
fuelectenerg-13681	143	22	eigenvalue	eigenvalue	NOUN
fuelectenerg-13681	143	23	of	of	ADP
fuelectenerg-13681	143	24	r.	r.	PROPN
fuelectenerg-13681	143	25	this	this	PRON
fuelectenerg-13681	143	26	provides	provide	VERB
fuelectenerg-13681	143	27	a	a	DET
fuelectenerg-13681	143	28	theoretical	theoretical	ADJ
fuelectenerg-13681	143	29	bound	bind	VERB
fuelectenerg-13681	143	30	on	on	ADP
fuelectenerg-13681	143	31	the	the	DET
fuelectenerg-13681	143	32	convergence	convergence	NOUN
fuelectenerg-13681	143	33	speed	speed	NOUN
fuelectenerg-13681	143	34	.	.	PUNCT
fuelectenerg-13681	144	1	recursive	recursive	ADJ
fuelectenerg-13681	144	2	least	least	ADJ
fuelectenerg-13681	144	3	squares	square	NOUN
fuelectenerg-13681	144	4	(	(	PUNCT
fuelectenerg-13681	144	5	rls	rls	PROPN
fuelectenerg-13681	144	6	):	):	PUNCT
fuelectenerg-13681	144	7	1	1	NUM
fuelectenerg-13681	144	8	1	1	NUM
fuelectenerg-13681	144	9	(	(	PUNCT
fuelectenerg-13681	144	10	1	1	NUM
fuelectenerg-13681	144	11	)	)	PUNCT
fuelectenerg-13681	144	12	(	(	PUNCT
fuelectenerg-13681	144	13	)	)	PUNCT
fuelectenerg-13681	144	14	(	(	PUNCT
fuelectenerg-13681	144	15	)	)	PUNCT
fuelectenerg-13681	144	16	1	1	NUM
fuelectenerg-13681	144	17	(	(	PUNCT
fuelectenerg-13681	144	18	)	)	PUNCT
fuelectenerg-13681	144	19	(	(	PUNCT
fuelectenerg-13681	144	20	1	1	X
fuelectenerg-13681	144	21	)	)	PUNCT
fuelectenerg-13681	144	22	(	(	PUNCT
fuelectenerg-13681	144	23	)	)	PUNCT
fuelectenerg-13681	144	24	h	h	PROPN
fuelectenerg-13681	144	25	n	n	CCONJ
fuelectenerg-13681	144	26	n	n	CCONJ
fuelectenerg-13681	144	27	n	n	CCONJ
fuelectenerg-13681	144	28	n	n	CCONJ
fuelectenerg-13681	144	29	n	n	CCONJ
fuelectenerg-13681	144	30	n	n	NUM
fuelectenerg-13681	144	31			X
fuelectenerg-13681	144	32			ADJ
fuelectenerg-13681	144	33	−	−	PROPN
fuelectenerg-13681	144	34	−	−	NOUN
fuelectenerg-13681	144	35	−	−	PROPN
fuelectenerg-13681	144	36	=	=	SYM
fuelectenerg-13681	145	1	+	+	NUM
fuelectenerg-13681	145	2	−	−	PROPN
fuelectenerg-13681	145	3	p	p	X
fuelectenerg-13681	145	4	x	x	X
fuelectenerg-13681	145	5	k	k	NOUN
fuelectenerg-13681	145	6	x	x	X
fuelectenerg-13681	145	7	p	p	X
fuelectenerg-13681	145	8	x	x	X
fuelectenerg-13681	145	9	(	(	PUNCT
fuelectenerg-13681	145	10	7	7	NUM
fuelectenerg-13681	145	11	)	)	PUNCT
fuelectenerg-13681	145	12	(	(	PUNCT
fuelectenerg-13681	145	13	)	)	PUNCT
fuelectenerg-13681	145	14	(	(	PUNCT
fuelectenerg-13681	145	15	)	)	PUNCT
fuelectenerg-13681	145	16	1	1	X
fuelectenerg-13681	145	17	)	)	PUNCT
fuelectenerg-13681	145	18	(	(	PUNCT
fuelectenerg-13681	145	19	)	)	PUNCT
fuelectenerg-13681	145	20	(	(	PUNCT
fuelectenerg-13681	145	21	hn	hn	PROPN
fuelectenerg-13681	145	22	d	d	PROPN
fuelectenerg-13681	145	23	n	n	CCONJ
fuelectenerg-13681	145	24	n	n	CCONJ
fuelectenerg-13681	145	25	n	n	ADV
fuelectenerg-13681	145	26	=	=	PUNCT
fuelectenerg-13681	145	27	−	−	PROPN
fuelectenerg-13681	145	28	−w	−w	ADV
fuelectenerg-13681	145	29	x	x	X
fuelectenerg-13681	145	30	(	(	PUNCT
fuelectenerg-13681	145	31	8)	8)	NUM
fuelectenerg-13681	145	32	*	*	PUNCT
fuelectenerg-13681	145	33	(	(	PUNCT
fuelectenerg-13681	145	34	)	)	PUNCT
fuelectenerg-13681	145	35	(	(	PUNCT
fuelectenerg-13681	145	36	1	1	X
fuelectenerg-13681	145	37	)	)	PUNCT
fuelectenerg-13681	145	38	(	(	PUNCT
fuelectenerg-13681	145	39	)	)	PUNCT
fuelectenerg-13681	145	40	(	(	PUNCT
fuelectenerg-13681	145	41	)	)	PUNCT
fuelectenerg-13681	145	42	n	n	CCONJ
fuelectenerg-13681	145	43	n	n	PROPN
fuelectenerg-13681	145	44	n	n	PRON
fuelectenerg-13681	145	45	n=	n=	NOUN
fuelectenerg-13681	145	46	−	−	PROPN
fuelectenerg-13681	146	1	+	+	PROPN
fuelectenerg-13681	146	2	w	w	PROPN
fuelectenerg-13681	146	3	w	w	PROPN
fuelectenerg-13681	146	4	k	k	PROPN
fuelectenerg-13681	146	5	(	(	PUNCT
fuelectenerg-13681	146	6	9	9	NUM
fuelectenerg-13681	146	7	)	)	SYM
fuelectenerg-13681	146	8	1	1	NUM
fuelectenerg-13681	146	9	1	1	NUM
fuelectenerg-13681	146	10	(	(	PUNCT
fuelectenerg-13681	146	11	)	)	PUNCT
fuelectenerg-13681	146	12	(	(	PUNCT
fuelectenerg-13681	146	13	1	1	X
fuelectenerg-13681	146	14	)	)	PUNCT
fuelectenerg-13681	146	15	(	(	PUNCT
fuelectenerg-13681	146	16	)	)	PUNCT
fuelectenerg-13681	146	17	(	(	PUNCT
fuelectenerg-13681	146	18	)	)	PUNCT
fuelectenerg-13681	146	19	(	(	PUNCT
fuelectenerg-13681	146	20	1)hn	1)hn	PROPN
fuelectenerg-13681	146	21	n	n	CCONJ
fuelectenerg-13681	146	22	n	n	CCONJ
fuelectenerg-13681	146	23	n	n	CCONJ
fuelectenerg-13681	146	24	n	n	PROPN
fuelectenerg-13681	146	25	−	−	NOUN
fuelectenerg-13681	146	26	−=	−=	NOUN
fuelectenerg-13681	146	27	−	−	PROPN
fuelectenerg-13681	146	28	−	−	NUM
fuelectenerg-13681	146	29	−p	−p	NOUN
fuelectenerg-13681	147	1	p	p	X
fuelectenerg-13681	147	2	k	k	PROPN
fuelectenerg-13681	147	3	x	x	X
fuelectenerg-13681	147	4	p	p	X
fuelectenerg-13681	147	5	(	(	PUNCT
fuelectenerg-13681	147	6	10	10	NUM
fuelectenerg-13681	147	7	)	)	PUNCT
fuelectenerg-13681	147	8	where	where	SCONJ
fuelectenerg-13681	147	9			ADJ
fuelectenerg-13681	147	10	is	be	AUX
fuelectenerg-13681	147	11	the	the	DET
fuelectenerg-13681	147	12	forgetting	forget	VERB
fuelectenerg-13681	147	13	factor	factor	NOUN
fuelectenerg-13681	147	14	,	,	PUNCT
fuelectenerg-13681	147	15	p(n	p(n	PROPN
fuelectenerg-13681	147	16	)	)	PUNCT
fuelectenerg-13681	147	17	is	be	AUX
fuelectenerg-13681	147	18	the	the	DET
fuelectenerg-13681	147	19	inverse	inverse	ADJ
fuelectenerg-13681	147	20	correlation	correlation	NOUN
fuelectenerg-13681	147	21	matrix	matrix	NOUN
fuelectenerg-13681	147	22	,	,	PUNCT
fuelectenerg-13681	147	23	and	and	CCONJ
fuelectenerg-13681	147	24	d(n	d(n	NOUN
fuelectenerg-13681	147	25	)	)	PUNCT
fuelectenerg-13681	147	26	is	be	AUX
fuelectenerg-13681	147	27	the	the	DET
fuelectenerg-13681	147	28	desired	desire	VERB
fuelectenerg-13681	147	29	signal	signal	NOUN
fuelectenerg-13681	147	30	.	.	PUNCT
fuelectenerg-13681	148	1	rls	rls	PROPN
fuelectenerg-13681	148	2	exhibits	exhibit	VERB
fuelectenerg-13681	148	3	significantly	significantly	ADV
fuelectenerg-13681	148	4	faster	fast	ADJ
fuelectenerg-13681	148	5	convergence	convergence	NOUN
fuelectenerg-13681	148	6	than	than	ADP
fuelectenerg-13681	148	7	lms	lm	NOUN
fuelectenerg-13681	148	8	,	,	PUNCT
fuelectenerg-13681	148	9	with	with	ADP
fuelectenerg-13681	148	10	a	a	DET
fuelectenerg-13681	148	11	convergence	convergence	NOUN
fuelectenerg-13681	148	12	rate	rate	NOUN
fuelectenerg-13681	148	13	that	that	PRON
fuelectenerg-13681	148	14	is	be	AUX
fuelectenerg-13681	148	15	approximately	approximately	ADV
fuelectenerg-13681	148	16	independent	independent	ADJ
fuelectenerg-13681	148	17	of	of	ADP
fuelectenerg-13681	148	18	the	the	DET
fuelectenerg-13681	148	19	eigenvalue	eigenvalue	ADJ
fuelectenerg-13681	148	20	spread	spread	NOUN
fuelectenerg-13681	148	21	of	of	ADP
fuelectenerg-13681	148	22	the	the	DET
fuelectenerg-13681	148	23	correlation	correlation	NOUN
fuelectenerg-13681	148	24	matrix	matrix	NOUN
fuelectenerg-13681	148	25	.	.	PUNCT
fuelectenerg-13681	149	1	its	its	PRON
fuelectenerg-13681	149	2	computational	computational	ADJ
fuelectenerg-13681	149	3	complexity	complexity	NOUN
fuelectenerg-13681	149	4	is	be	AUX
fuelectenerg-13681	149	5	2	2	NUM
fuelectenerg-13681	149	6	(	(	PUNCT
fuelectenerg-13681	149	7	)	)	PUNCT
fuelectenerg-13681	149	8	to	to	ADP
fuelectenerg-13681	149	9	n	n	PROPN
fuelectenerg-13681	149	10	per	per	ADP
fuelectenerg-13681	149	11	iteration	iteration	NOUN
fuelectenerg-13681	149	12	,	,	PUNCT
fuelectenerg-13681	149	13	compared	compare	VERB
fuelectenerg-13681	149	14	to	to	ADP
fuelectenerg-13681	149	15	(	(	PUNCT
fuelectenerg-13681	149	16	)	)	PUNCT
fuelectenerg-13681	149	17	to	to	ADP
fuelectenerg-13681	149	18	n	n	NOUN
fuelectenerg-13681	149	19	or	or	CCONJ
fuelectenerg-13681	149	20	lms	lm	NOUN
fuelectenerg-13681	149	21	.	.	PUNCT
fuelectenerg-13681	150	1	minimum	minimum	ADJ
fuelectenerg-13681	150	2	variance	variance	NOUN
fuelectenerg-13681	150	3	distortionless	distortionless	NOUN
fuelectenerg-13681	150	4	response	response	NOUN
fuelectenerg-13681	150	5	(	(	PUNCT
fuelectenerg-13681	150	6	mvdr	mvdr	NOUN
fuelectenerg-13681	150	7	):	):	PUNCT
fuelectenerg-13681	150	8	1	1	NUM
fuelectenerg-13681	150	9	1	1	NUM
fuelectenerg-13681	150	10	(	(	PUNCT
fuelectenerg-13681	150	11	)	)	PUNCT
fuelectenerg-13681	150	12	(	(	PUNCT
fuelectenerg-13681	150	13	)	)	PUNCT
fuelectenerg-13681	150	14	(	(	PUNCT
fuelectenerg-13681	150	15	)	)	PUNCT
fuelectenerg-13681	150	16	d	d	NOUN
fuelectenerg-13681	150	17	mvdr	mvdr	VERB
fuelectenerg-13681	150	18	h	h	NOUN
fuelectenerg-13681	151	1	d	d	PROPN
fuelectenerg-13681	151	2	d	d	X
fuelectenerg-13681	151	3			X
fuelectenerg-13681	151	4			X
fuelectenerg-13681	151	5			PROPN
fuelectenerg-13681	152	1	−	−	PROPN
fuelectenerg-13681	153	1	−	−	PROPN
fuelectenerg-13681	153	2	=	=	SYM
fuelectenerg-13681	153	3	r	r	NOUN
fuelectenerg-13681	153	4	a	a	DET
fuelectenerg-13681	153	5	w	w	NOUN
fuelectenerg-13681	153	6	a	a	DET
fuelectenerg-13681	153	7	r	r	NOUN
fuelectenerg-13681	153	8	a	a	DET
fuelectenerg-13681	153	9	(	(	PUNCT
fuelectenerg-13681	153	10	11	11	NUM
fuelectenerg-13681	153	11	)	)	PUNCT
fuelectenerg-13681	153	12	where	where	SCONJ
fuelectenerg-13681	153	13	r	r	NOUN
fuelectenerg-13681	153	14	is	be	AUX
fuelectenerg-13681	153	15	the	the	DET
fuelectenerg-13681	153	16	interference	interference	NOUN
fuelectenerg-13681	153	17	-	-	PUNCT
fuelectenerg-13681	153	18	plus	plus	CCONJ
fuelectenerg-13681	153	19	-	-	PUNCT
fuelectenerg-13681	153	20	noise	noise	NOUN
fuelectenerg-13681	153	21	covariance	covariance	NOUN
fuelectenerg-13681	153	22	matrix	matrix	NOUN
fuelectenerg-13681	153	23	and	and	CCONJ
fuelectenerg-13681	153	24	a(d	a(d	NOUN
fuelectenerg-13681	153	25	)	)	PUNCT
fuelectenerg-13681	153	26	is	be	AUX
fuelectenerg-13681	153	27	the	the	DET
fuelectenerg-13681	153	28	array	array	NOUN
fuelectenerg-13681	153	29	response	response	NOUN
fuelectenerg-13681	153	30	vector	vector	NOUN
fuelectenerg-13681	153	31	in	in	ADP
fuelectenerg-13681	153	32	the	the	DET
fuelectenerg-13681	153	33	desired	desire	VERB
fuelectenerg-13681	153	34	direction	direction	NOUN
fuelectenerg-13681	153	35	.	.	PUNCT
fuelectenerg-13681	154	1	the	the	DET
fuelectenerg-13681	154	2	mvdr	mvdr	NOUN
fuelectenerg-13681	154	3	beamformer	beamformer	NOUN
fuelectenerg-13681	154	4	is	be	AUX
fuelectenerg-13681	154	5	optimal	optimal	ADJ
fuelectenerg-13681	154	6	in	in	ADP
fuelectenerg-13681	154	7	the	the	DET
fuelectenerg-13681	154	8	sense	sense	NOUN
fuelectenerg-13681	154	9	that	that	SCONJ
fuelectenerg-13681	154	10	it	it	PRON
fuelectenerg-13681	154	11	minimizes	minimize	VERB
fuelectenerg-13681	154	12	the	the	DET
fuelectenerg-13681	154	13	output	output	NOUN
fuelectenerg-13681	154	14	power	power	NOUN
fuelectenerg-13681	154	15	subject	subject	ADJ
fuelectenerg-13681	154	16	to	to	ADP
fuelectenerg-13681	154	17	maintaining	maintain	VERB
fuelectenerg-13681	154	18	unity	unity	NOUN
fuelectenerg-13681	154	19	gain	gain	NOUN
fuelectenerg-13681	154	20	in	in	ADP
fuelectenerg-13681	154	21	the	the	DET
fuelectenerg-13681	154	22	desired	desire	VERB
fuelectenerg-13681	154	23	direction	direction	NOUN
fuelectenerg-13681	154	24	.	.	PUNCT
fuelectenerg-13681	155	1	its	its	PRON
fuelectenerg-13681	155	2	sinr	sinr	NOUN
fuelectenerg-13681	155	3	is	be	AUX
fuelectenerg-13681	155	4	bounded	bound	VERB
fuelectenerg-13681	155	5	by	by	ADP
fuelectenerg-13681	155	6	:	:	PUNCT
fuelectenerg-13681	155	7	204	204	NUM
fuelectenerg-13681	155	8	m.	m.	NOUN
fuelectenerg-13681	155	9	mohamed	mohamed	PROPN
fuelectenerg-13681	155	10	,	,	PUNCT
fuelectenerg-13681	155	11	f.	f.	PROPN
fuelectenerg-13681	155	12	aljuaid	aljuaid	PROPN
fuelectenerg-13681	155	13	,	,	PUNCT
fuelectenerg-13681	155	14	mhd	mhd	PROPN
fuelectenerg-13681	155	15	w.	w.	PROPN
fuelectenerg-13681	155	16	koubeisi	koubeisi	VERB
fuelectenerg-13681	155	17	1	1	NUM
fuelectenerg-13681	155	18	2	2	NUM
fuelectenerg-13681	155	19	sinr	sinr	NOUN
fuelectenerg-13681	155	20	(	(	PUNCT
fuelectenerg-13681	155	21	)	)	PUNCT
fuelectenerg-13681	155	22	(	(	PUNCT
fuelectenerg-13681	155	23	)	)	PUNCT
fuelectenerg-13681	155	24	hs	hs	PROPN
fuelectenerg-13681	155	25	mvdr	mvdr	VERB
fuelectenerg-13681	156	1	d	d	NOUN
fuelectenerg-13681	156	2	i	i	NOUN
fuelectenerg-13681	157	1	d	d	X
fuelectenerg-13681	157	2	p	p	X
fuelectenerg-13681	157	3			PROPN
fuelectenerg-13681	157	4			X
fuelectenerg-13681	157	5			X
fuelectenerg-13681	157	6	−	−	X
fuelectenerg-13681	157	7	a	a	PROPN
fuelectenerg-13681	157	8	r	r	NOUN
fuelectenerg-13681	157	9	a	a	PRON
fuelectenerg-13681	157	10	(	(	PUNCT
fuelectenerg-13681	157	11	12	12	NUM
fuelectenerg-13681	157	12	)	)	PUNCT
fuelectenerg-13681	157	13	where	where	SCONJ
fuelectenerg-13681	157	14	ps	ps	PROPN
fuelectenerg-13681	157	15	is	be	AUX
fuelectenerg-13681	157	16	the	the	DET
fuelectenerg-13681	157	17	signal	signal	ADJ
fuelectenerg-13681	157	18	power	power	NOUN
fuelectenerg-13681	157	19	and	and	CCONJ
fuelectenerg-13681	157	20	ri	ri	PROPN
fuelectenerg-13681	157	21	is	be	AUX
fuelectenerg-13681	157	22	the	the	DET
fuelectenerg-13681	157	23	interference	interference	NOUN
fuelectenerg-13681	157	24	covariance	covariance	NOUN
fuelectenerg-13681	157	25	matrix	matrix	NOUN
fuelectenerg-13681	157	26	.	.	PUNCT
fuelectenerg-13681	158	1	second	second	ADJ
fuelectenerg-13681	158	2	-	-	PUNCT
fuelectenerg-13681	158	3	order	order	NOUN
fuelectenerg-13681	158	4	cone	cone	NOUN
fuelectenerg-13681	158	5	programming	programming	NOUN
fuelectenerg-13681	158	6	(	(	PUNCT
fuelectenerg-13681	158	7	socp	socp	ADJ
fuelectenerg-13681	158	8	):	):	PUNCT
fuelectenerg-13681	158	9	2	2	NUM
fuelectenerg-13681	158	10	2	2	NUM
fuelectenerg-13681	158	11	min2	min2	NOUN
fuelectenerg-13681	158	12	2	2	NUM
fuelectenerg-13681	158	13	|	|	ADV
fuelectenerg-13681	158	14	|	|	ADV
fuelectenerg-13681	158	15	min	min	NOUN
fuelectenerg-13681	159	1	|	|	ADV
fuelectenerg-13681	159	2	|	|	ADV
fuelectenerg-13681	159	3	    	    	SPACE
fuelectenerg-13681	159	4	s.t	s.t	PROPN
fuelectenerg-13681	159	5	.	.	PROPN
fuelectenerg-13681	159	6	     	     	SPACE
fuelectenerg-13681	159	7	‍|	‍|	PROPN
fuelectenerg-13681	160	1	|	|	NOUN
fuelectenerg-13681	161	1	d	d	NOUN
fuelectenerg-13681	161	2	i	i	PRON
fuelectenerg-13681	162	1	i	i	VERB
fuelectenerg-13681	163	1	d	d	X
fuelectenerg-13681	163	2			NUM
fuelectenerg-13681	163	3			PROPN
fuelectenerg-13681	163	4			PROPN
fuelectenerg-13681	163	5			NUM
fuelectenerg-13681	164	1	+	+	NOUN
fuelectenerg-13681	164	2	w	w	NOUN
fuelectenerg-13681	164	3	h	h	NOUN
fuelectenerg-13681	164	4	w	w	PROPN
fuelectenerg-13681	164	5	w	w	PROPN
fuelectenerg-13681	164	6	h	h	PROPN
fuelectenerg-13681	164	7	w	w	PROPN
fuelectenerg-13681	164	8	(	(	PUNCT
fuelectenerg-13681	164	9	13	13	NUM
fuelectenerg-13681	164	10	)	)	PUNCT
fuelectenerg-13681	164	11	where	where	SCONJ
fuelectenerg-13681	164	12	min	min	NOUN
fuelectenerg-13681	164	13	is	be	AUX
fuelectenerg-13681	164	14	the	the	DET
fuelectenerg-13681	164	15	minimum	minimum	NOUN
fuelectenerg-13681	164	16	required	require	VERB
fuelectenerg-13681	164	17	sinr	sinr	NOUN
fuelectenerg-13681	164	18	.	.	PUNCT
fuelectenerg-13681	165	1	this	this	DET
fuelectenerg-13681	165	2	formulation	formulation	NOUN
fuelectenerg-13681	165	3	is	be	AUX
fuelectenerg-13681	165	4	a	a	DET
fuelectenerg-13681	165	5	convex	convex	ADJ
fuelectenerg-13681	165	6	optimization	optimization	NOUN
fuelectenerg-13681	165	7	problem	problem	NOUN
fuelectenerg-13681	165	8	that	that	PRON
fuelectenerg-13681	165	9	can	can	AUX
fuelectenerg-13681	165	10	be	be	AUX
fuelectenerg-13681	165	11	solved	solve	VERB
fuelectenerg-13681	165	12	using	use	VERB
fuelectenerg-13681	165	13	interior	interior	ADJ
fuelectenerg-13681	165	14	-	-	PUNCT
fuelectenerg-13681	165	15	point	point	NOUN
fuelectenerg-13681	165	16	methods	method	NOUN
fuelectenerg-13681	165	17	with	with	ADP
fuelectenerg-13681	165	18	polynomial	polynomial	ADJ
fuelectenerg-13681	165	19	time	time	NOUN
fuelectenerg-13681	165	20	complexity	complexity	NOUN
fuelectenerg-13681	165	21	.	.	PUNCT
fuelectenerg-13681	166	1	for	for	ADP
fuelectenerg-13681	166	2	an	an	DET
fuelectenerg-13681	166	3	nt	not	PART
fuelectenerg-13681	166	4			PROPN
fuelectenerg-13681	166	5	nr	nr	PROPN
fuelectenerg-13681	166	6	mimo	mimo	PROPN
fuelectenerg-13681	166	7	system	system	PROPN
fuelectenerg-13681	166	8	,	,	PUNCT
fuelectenerg-13681	166	9	the	the	DET
fuelectenerg-13681	166	10	computational	computational	ADJ
fuelectenerg-13681	166	11	complexity	complexity	NOUN
fuelectenerg-13681	166	12	is	be	AUX
fuelectenerg-13681	166	13	approximately	approximately	ADV
fuelectenerg-13681	166	14	3	3	NUM
fuelectenerg-13681	166	15	3	3	NUM
fuelectenerg-13681	166	16	(	(	PUNCT
fuelectenerg-13681	166	17	)	)	PUNCT
fuelectenerg-13681	166	18	t	t	PROPN
fuelectenerg-13681	166	19	ro	ro	NOUN
fuelectenerg-13681	166	20	n	n	CCONJ
fuelectenerg-13681	166	21	n	n	PROPN
fuelectenerg-13681	166	22	or	or	CCONJ
fuelectenerg-13681	166	23	a	a	DET
fuelectenerg-13681	166	24	typical	typical	ADJ
fuelectenerg-13681	166	25	interior	interior	ADJ
fuelectenerg-13681	166	26	-	-	PUNCT
fuelectenerg-13681	166	27	point	point	NOUN
fuelectenerg-13681	166	28	solver	solver	NOUN
fuelectenerg-13681	166	29	.	.	PUNCT
fuelectenerg-13681	167	1	3.3	3.3	NUM
fuelectenerg-13681	167	2	.	.	PUNCT
fuelectenerg-13681	168	1	deep	deep	ADJ
fuelectenerg-13681	168	2	learning	learning	NOUN
fuelectenerg-13681	168	3	approaches	approach	NOUN
fuelectenerg-13681	168	4	we	we	PRON
fuelectenerg-13681	168	5	implement	implement	VERB
fuelectenerg-13681	168	6	four	four	NUM
fuelectenerg-13681	168	7	deep	deep	ADJ
fuelectenerg-13681	168	8	learning	learning	NOUN
fuelectenerg-13681	168	9	architectures	architecture	NOUN
fuelectenerg-13681	168	10	,	,	PUNCT
fuelectenerg-13681	168	11	each	each	PRON
fuelectenerg-13681	168	12	designed	design	VERB
fuelectenerg-13681	168	13	to	to	PART
fuelectenerg-13681	168	14	address	address	VERB
fuelectenerg-13681	168	15	specific	specific	ADJ
fuelectenerg-13681	168	16	aspects	aspect	NOUN
fuelectenerg-13681	168	17	of	of	ADP
fuelectenerg-13681	168	18	the	the	DET
fuelectenerg-13681	168	19	beamforming	beamforming	NOUN
fuelectenerg-13681	168	20	problem	problem	NOUN
fuelectenerg-13681	168	21	:	:	PUNCT
fuelectenerg-13681	168	22	the	the	DET
fuelectenerg-13681	168	23	architecture	architecture	NOUN
fuelectenerg-13681	168	24	can	can	AUX
fuelectenerg-13681	168	25	be	be	AUX
fuelectenerg-13681	168	26	formally	formally	ADV
fuelectenerg-13681	168	27	described	describe	VERB
fuelectenerg-13681	168	28	as	as	ADP
fuelectenerg-13681	168	29	:	:	PUNCT
fuelectenerg-13681	168	30	(	(	PUNCT
fuelectenerg-13681	168	31	)	)	SYM
fuelectenerg-13681	168	32	1	1	NUM
fuelectenerg-13681	168	33	1	1	NUM
fuelectenerg-13681	168	34	in	in	ADP
fuelectenerg-13681	168	35	1relu	1relu	NUM
fuelectenerg-13681	168	36	bn	bn	NOUN
fuelectenerg-13681	168	37	(	(	PUNCT
fuelectenerg-13681	168	38	)	)	PUNCT
fuelectenerg-13681	168	39	=	=	PUNCT
fuelectenerg-13681	169	1	+	+	PUNCT
fuelectenerg-13681	169	2	h	h	NOUN
fuelectenerg-13681	169	3	wh	wh	NOUN
fuelectenerg-13681	169	4	b	b	PROPN
fuelectenerg-13681	169	5	(	(	PUNCT
fuelectenerg-13681	169	6	14	14	NUM
fuelectenerg-13681	169	7	)	)	PUNCT
fuelectenerg-13681	169	8	(	(	PUNCT
fuelectenerg-13681	169	9	)	)	SYM
fuelectenerg-13681	169	10	2	2	NUM
fuelectenerg-13681	169	11	2	2	NUM
fuelectenerg-13681	169	12	1	1	NUM
fuelectenerg-13681	169	13	2relu	2relu	NUM
fuelectenerg-13681	169	14	bn	bn	NOUN
fuelectenerg-13681	169	15	(	(	PUNCT
fuelectenerg-13681	169	16	)	)	PUNCT
fuelectenerg-13681	169	17	=	=	PUNCT
fuelectenerg-13681	170	1	+	+	NOUN
fuelectenerg-13681	170	2	h	h	PROPN
fuelectenerg-13681	170	3	w	w	PROPN
fuelectenerg-13681	170	4	h	h	PROPN
fuelectenerg-13681	170	5	b	b	PROPN
fuelectenerg-13681	170	6	(	(	PUNCT
fuelectenerg-13681	170	7	15	15	NUM
fuelectenerg-13681	170	8	)	)	PUNCT
fuelectenerg-13681	170	9	(	(	PUNCT
fuelectenerg-13681	170	10	)	)	SYM
fuelectenerg-13681	170	11	3	3	NUM
fuelectenerg-13681	170	12	3	3	NUM
fuelectenerg-13681	170	13	2	2	NUM
fuelectenerg-13681	170	14	3relu	3relu	NUM
fuelectenerg-13681	170	15	bn	bn	NUM
fuelectenerg-13681	170	16	(	(	PUNCT
fuelectenerg-13681	170	17	)	)	PUNCT
fuelectenerg-13681	170	18	=	=	PUNCT
fuelectenerg-13681	171	1	+	+	NOUN
fuelectenerg-13681	171	2	h	h	PROPN
fuelectenerg-13681	171	3	w	w	PROPN
fuelectenerg-13681	171	4	h	h	PROPN
fuelectenerg-13681	171	5	b	b	PROPN
fuelectenerg-13681	171	6	(	(	PUNCT
fuelectenerg-13681	171	7	16	16	NUM
fuelectenerg-13681	171	8	)	)	PUNCT
fuelectenerg-13681	171	9	(	(	PUNCT
fuelectenerg-13681	171	10	)	)	PUNCT
fuelectenerg-13681	171	11	4	4	NUM
fuelectenerg-13681	171	12	4	4	NUM
fuelectenerg-13681	171	13	3	3	NUM
fuelectenerg-13681	171	14	4relu	4relu	NUM
fuelectenerg-13681	171	15	bn	bn	NOUN
fuelectenerg-13681	171	16	(	(	PUNCT
fuelectenerg-13681	171	17	)	)	PUNCT
fuelectenerg-13681	171	18	=	=	PUNCT
fuelectenerg-13681	172	1	+	+	NOUN
fuelectenerg-13681	172	2	h	h	PROPN
fuelectenerg-13681	172	3	w	w	PROPN
fuelectenerg-13681	172	4	h	h	PROPN
fuelectenerg-13681	172	5	b	b	PROPN
fuelectenerg-13681	172	6	(	(	PUNCT
fuelectenerg-13681	172	7	17	17	NUM
fuelectenerg-13681	172	8	)	)	PUNCT
fuelectenerg-13681	172	9	out	out	ADV
fuelectenerg-13681	172	10	5	5	NUM
fuelectenerg-13681	172	11	4	4	NUM
fuelectenerg-13681	172	12	5=	5=	NUM
fuelectenerg-13681	172	13	+	+	PROPN
fuelectenerg-13681	172	14	h	h	PROPN
fuelectenerg-13681	172	15	w	w	PROPN
fuelectenerg-13681	172	16	h	h	PROPN
fuelectenerg-13681	172	17	b	b	PROPN
fuelectenerg-13681	172	18	(	(	PUNCT
fuelectenerg-13681	172	19	18	18	NUM
fuelectenerg-13681	172	20	)	)	PUNCT
fuelectenerg-13681	172	21	where	where	SCONJ
fuelectenerg-13681	172	22	hin	hin	PROPN
fuelectenerg-13681	172	23	is	be	AUX
fuelectenerg-13681	172	24	the	the	DET
fuelectenerg-13681	172	25	flattened	flatten	VERB
fuelectenerg-13681	172	26	channel	channel	NOUN
fuelectenerg-13681	172	27	matrix	matrix	NOUN
fuelectenerg-13681	172	28	,	,	PUNCT
fuelectenerg-13681	172	29	hout	hout	PROPN
fuelectenerg-13681	172	30	is	be	AUX
fuelectenerg-13681	172	31	the	the	DET
fuelectenerg-13681	172	32	predicted	predict	VERB
fuelectenerg-13681	172	33	beamforming	beamforme	VERB
fuelectenerg-13681	172	34	weights	weight	NOUN
fuelectenerg-13681	172	35	,	,	PUNCT
fuelectenerg-13681	172	36	wi	wi	PROPN
fuelectenerg-13681	172	37	and	and	CCONJ
fuelectenerg-13681	172	38	bi	bi	NOUN
fuelectenerg-13681	172	39	are	be	AUX
fuelectenerg-13681	172	40	the	the	DET
fuelectenerg-13681	172	41	weights	weight	NOUN
fuelectenerg-13681	172	42	and	and	CCONJ
fuelectenerg-13681	172	43	biases	bias	NOUN
fuelectenerg-13681	172	44	of	of	ADP
fuelectenerg-13681	172	45	layer	layer	NOUN
fuelectenerg-13681	172	46	i	i	PRON
fuelectenerg-13681	172	47	and	and	CCONJ
fuelectenerg-13681	173	1	bn	bn	NOUN
fuelectenerg-13681	173	2	denotes	denote	NOUN
fuelectenerg-13681	173	3	batch	batch	VERB
fuelectenerg-13681	173	4	normalization	normalization	NOUN
fuelectenerg-13681	173	5	.	.	PUNCT
fuelectenerg-13681	174	1	the	the	DET
fuelectenerg-13681	174	2	convolutional	convolutional	ADJ
fuelectenerg-13681	174	3	layers	layer	NOUN
fuelectenerg-13681	174	4	use	use	VERB
fuelectenerg-13681	174	5	3	3	NUM
fuelectenerg-13681	174	6			NOUN
fuelectenerg-13681	174	7	3	3	NUM
fuelectenerg-13681	174	8	filters	filter	NOUN
fuelectenerg-13681	174	9	with	with	ADP
fuelectenerg-13681	174	10	stride	stride	NOUN
fuelectenerg-13681	174	11	1	1	NUM
fuelectenerg-13681	174	12	and	and	CCONJ
fuelectenerg-13681	174	13	’	'	PUNCT
fuelectenerg-13681	174	14	same	same	ADJ
fuelectenerg-13681	174	15	’	'	PUNCT
fuelectenerg-13681	174	16	padding	padding	NOUN
fuelectenerg-13681	174	17	,	,	PUNCT
fuelectenerg-13681	174	18	followed	follow	VERB
fuelectenerg-13681	174	19	by	by	ADP
fuelectenerg-13681	174	20	2	2	NUM
fuelectenerg-13681	174	21			NOUN
fuelectenerg-13681	174	22	2	2	NUM
fuelectenerg-13681	174	23	ax	ax	NOUN
fuelectenerg-13681	174	24	-	-	PUNCT
fuelectenerg-13681	174	25	pooling	pooling	NOUN
fuelectenerg-13681	174	26	.	.	PUNCT
fuelectenerg-13681	175	1	the	the	DET
fuelectenerg-13681	175	2	mathematical	mathematical	ADJ
fuelectenerg-13681	175	3	formulation	formulation	NOUN
fuelectenerg-13681	175	4	is	be	AUX
fuelectenerg-13681	175	5	:	:	PUNCT
fuelectenerg-13681	175	6	(	(	PUNCT
fuelectenerg-13681	175	7	)	)	SYM
fuelectenerg-13681	175	8	1	1	NUM
fuelectenerg-13681	175	9	3	3	NUM
fuelectenerg-13681	175	10	3,64	3,64	NUM
fuelectenerg-13681	175	11	inmaxpool	inmaxpool	NOUN
fuelectenerg-13681	175	12	relu(conv	relu(conv	PROPN
fuelectenerg-13681	175	13	)	)	PUNCT
fuelectenerg-13681	175	14	(	(	PUNCT
fuelectenerg-13681	175	15	)	)	PUNCT
fuelectenerg-13681	175	16	=h	=h	PROPN
fuelectenerg-13681	175	17	h	h	NOUN
fuelectenerg-13681	175	18	(	(	PUNCT
fuelectenerg-13681	175	19	19	19	NUM
fuelectenerg-13681	175	20	)	)	PUNCT
fuelectenerg-13681	175	21	(	(	PUNCT
fuelectenerg-13681	175	22	)	)	SYM
fuelectenerg-13681	175	23	2	2	NUM
fuelectenerg-13681	175	24	3	3	NUM
fuelectenerg-13681	175	25	3,32	3,32	NUM
fuelectenerg-13681	175	26	1maxpool	1maxpool	PROPN
fuelectenerg-13681	175	27	relu(conv	relu(conv	PROPN
fuelectenerg-13681	175	28	)	)	PUNCT
fuelectenerg-13681	175	29	(	(	PUNCT
fuelectenerg-13681	175	30	)	)	PUNCT
fuelectenerg-13681	175	31	=h	=h	PROPN
fuelectenerg-13681	175	32	h	h	NOUN
fuelectenerg-13681	175	33	(	(	PUNCT
fuelectenerg-13681	175	34	20	20	NUM
fuelectenerg-13681	175	35	)	)	PUNCT
fuelectenerg-13681	175	36	(	(	PUNCT
fuelectenerg-13681	175	37	)	)	PUNCT
fuelectenerg-13681	175	38	3	3	NUM
fuelectenerg-13681	175	39	3	3	NUM
fuelectenerg-13681	175	40	3,16	3,16	PROPN
fuelectenerg-13681	175	41	2maxpool	2maxpool	NUM
fuelectenerg-13681	175	42	relu(conv	relu(conv	PROPN
fuelectenerg-13681	175	43	)	)	PUNCT
fuelectenerg-13681	175	44	(	(	PUNCT
fuelectenerg-13681	175	45	)	)	PUNCT
fuelectenerg-13681	175	46	=h	=h	PROPN
fuelectenerg-13681	175	47	h	h	NOUN
fuelectenerg-13681	175	48	(	(	PUNCT
fuelectenerg-13681	175	49	21	21	NUM
fuelectenerg-13681	175	50	)	)	PUNCT
fuelectenerg-13681	175	51	(	(	PUNCT
fuelectenerg-13681	175	52	)	)	PUNCT
fuelectenerg-13681	175	53	4	4	NUM
fuelectenerg-13681	175	54	4	4	NUM
fuelectenerg-13681	175	55	3	3	NUM
fuelectenerg-13681	175	56	4relu	4relu	NUM
fuelectenerg-13681	175	57	flatten	flatten	ADJ
fuelectenerg-13681	175	58	(	(	PUNCT
fuelectenerg-13681	175	59	)	)	PUNCT
fuelectenerg-13681	175	60	=	=	PUNCT
fuelectenerg-13681	176	1	+	+	NOUN
fuelectenerg-13681	176	2	h	h	PROPN
fuelectenerg-13681	176	3	w	w	PROPN
fuelectenerg-13681	176	4	h	h	PROPN
fuelectenerg-13681	176	5	b	b	PROPN
fuelectenerg-13681	176	6	(	(	PUNCT
fuelectenerg-13681	176	7	22	22	NUM
fuelectenerg-13681	176	8	)	)	PUNCT
fuelectenerg-13681	176	9	out	out	ADV
fuelectenerg-13681	176	10	5	5	NUM
fuelectenerg-13681	176	11	4	4	NUM
fuelectenerg-13681	176	12	5=	5=	NUM
fuelectenerg-13681	176	13	+	+	PROPN
fuelectenerg-13681	176	14	h	h	PROPN
fuelectenerg-13681	176	15	w	w	PROPN
fuelectenerg-13681	176	16	h	h	PROPN
fuelectenerg-13681	176	17	b	b	PROPN
fuelectenerg-13681	176	18	(	(	PUNCT
fuelectenerg-13681	176	19	23	23	NUM
fuelectenerg-13681	176	20	)	)	PUNCT
fuelectenerg-13681	176	21	where	where	SCONJ
fuelectenerg-13681	176	22	hin	hin	PROPN
fuelectenerg-13681	176	23	is	be	AUX
fuelectenerg-13681	176	24	the	the	DET
fuelectenerg-13681	176	25	input	input	NOUN
fuelectenerg-13681	176	26	channel	channel	NOUN
fuelectenerg-13681	176	27	matrix	matrix	NOUN
fuelectenerg-13681	176	28	and	and	CCONJ
fuelectenerg-13681	176	29	convkk	convkk	PROPN
fuelectenerg-13681	176	30	,	,	PUNCT
fuelectenerg-13681	176	31	f	f	PROPN
fuelectenerg-13681	176	32	denotes	denote	VERB
fuelectenerg-13681	176	33	a	a	DET
fuelectenerg-13681	176	34	convolutional	convolutional	ADJ
fuelectenerg-13681	176	35	layer	layer	NOUN
fuelectenerg-13681	176	36	with	with	ADP
fuelectenerg-13681	176	37	k	k	PROPN
fuelectenerg-13681	176	38			PROPN
fuelectenerg-13681	176	39	k	k	PROPN
fuelectenerg-13681	176	40	filters	filter	NOUN
fuelectenerg-13681	176	41	and	and	CCONJ
fuelectenerg-13681	176	42	f	f	NUM
fuelectenerg-13681	176	43	feature	feature	NOUN
fuelectenerg-13681	176	44	maps	map	NOUN
fuelectenerg-13681	176	45	.	.	PUNCT
fuelectenerg-13681	177	1	the	the	DET
fuelectenerg-13681	177	2	lstm	lstm	PROPN
fuelectenerg-13681	177	3	cell	cell	NOUN
fuelectenerg-13681	177	4	at	at	ADP
fuelectenerg-13681	177	5	time	time	NOUN
fuelectenerg-13681	177	6	step	step	NOUN
fuelectenerg-13681	177	7	t	t	NOUN
fuelectenerg-13681	177	8	and	and	CCONJ
fuelectenerg-13681	177	9	layer	layer	NOUN
fuelectenerg-13681	177	10	l	l	NOUN
fuelectenerg-13681	177	11	is	be	AUX
fuelectenerg-13681	177	12	described	describe	VERB
fuelectenerg-13681	177	13	by	by	ADP
fuelectenerg-13681	177	14	:	:	PUNCT
fuelectenerg-13681	177	15	1	1	NUM
fuelectenerg-13681	177	16	1	1	NUM
fuelectenerg-13681	177	17	(	(	PUNCT
fuelectenerg-13681	177	18	)	)	PUNCT
fuelectenerg-13681	177	19	l	l	NOUN
fuelectenerg-13681	177	20	l	l	NOUN
fuelectenerg-13681	177	21	l	l	NOUN
fuelectenerg-13681	177	22	l	l	X
fuelectenerg-13681	178	1	l	l	X
fuelectenerg-13681	178	2	l	l	NOUN
fuelectenerg-13681	178	3	t	t	NOUN
fuelectenerg-13681	179	1	f	f	PROPN
fuelectenerg-13681	179	2	t	t	PROPN
fuelectenerg-13681	179	3	f	f	PROPN
fuelectenerg-13681	179	4	t	t	PROPN
fuelectenerg-13681	179	5	f	f	ADV
fuelectenerg-13681	179	6	−	−	PROPN
fuelectenerg-13681	180	1	−=	−=	NOUN
fuelectenerg-13681	180	2	+	+	X
fuelectenerg-13681	181	1	+	+	NOUN
fuelectenerg-13681	181	2	f	f	PROPN
fuelectenerg-13681	181	3	w	w	NOUN
fuelectenerg-13681	181	4	h	h	NOUN
fuelectenerg-13681	181	5	u	u	NOUN
fuelectenerg-13681	181	6	h	h	PROPN
fuelectenerg-13681	181	7	b	b	PROPN
fuelectenerg-13681	181	8	(	(	PUNCT
fuelectenerg-13681	181	9	24	24	NUM
fuelectenerg-13681	181	10	)	)	SYM
fuelectenerg-13681	181	11	1	1	NUM
fuelectenerg-13681	181	12	1	1	NUM
fuelectenerg-13681	181	13	(	(	PUNCT
fuelectenerg-13681	181	14	)	)	PUNCT
fuelectenerg-13681	181	15	l	l	NOUN
fuelectenerg-13681	181	16	l	l	NOUN
fuelectenerg-13681	181	17	l	l	NOUN
fuelectenerg-13681	181	18	l	l	X
fuelectenerg-13681	181	19	l	l	X
fuelectenerg-13681	181	20	l	l	NOUN
fuelectenerg-13681	181	21	t	t	NOUN
fuelectenerg-13681	182	1	i	i	PRON
fuelectenerg-13681	182	2	t	t	X
fuelectenerg-13681	183	1	i	i	PRON
fuelectenerg-13681	183	2	t	t	PROPN
fuelectenerg-13681	183	3	i	i	PROPN
fuelectenerg-13681	183	4	−	−	PROPN
fuelectenerg-13681	183	5	−=	−=	NOUN
fuelectenerg-13681	183	6	+	+	PROPN
fuelectenerg-13681	184	1	+	+	ADV
fuelectenerg-13681	184	2	i	i	PROPN
fuelectenerg-13681	184	3	w	w	VERB
fuelectenerg-13681	184	4	h	h	NOUN
fuelectenerg-13681	184	5	u	u	NOUN
fuelectenerg-13681	184	6	h	h	PROPN
fuelectenerg-13681	184	7	b	b	PROPN
fuelectenerg-13681	184	8	(	(	PUNCT
fuelectenerg-13681	184	9	25	25	NUM
fuelectenerg-13681	184	10	)	)	PUNCT
fuelectenerg-13681	184	11	deep	deep	ADJ
fuelectenerg-13681	184	12	learning	learning	NOUN
fuelectenerg-13681	184	13	and	and	CCONJ
fuelectenerg-13681	184	14	traditional	traditional	ADJ
fuelectenerg-13681	184	15	algorithms	algorithm	NOUN
fuelectenerg-13681	184	16	for	for	ADP
fuelectenerg-13681	184	17	mimo	mimo	PROPN
fuelectenerg-13681	184	18	beamforming	beamforme	VERB
fuelectenerg-13681	184	19	205	205	NUM
fuelectenerg-13681	184	20	1	1	NUM
fuelectenerg-13681	184	21	1	1	NUM
fuelectenerg-13681	184	22	(	(	PUNCT
fuelectenerg-13681	184	23	)	)	PUNCT
fuelectenerg-13681	184	24	l	l	NOUN
fuelectenerg-13681	184	25	l	l	NOUN
fuelectenerg-13681	184	26	l	l	NOUN
fuelectenerg-13681	184	27	l	l	X
fuelectenerg-13681	184	28	l	l	X
fuelectenerg-13681	184	29	l	l	NOUN
fuelectenerg-13681	185	1	t	t	NOUN
fuelectenerg-13681	185	2	o	o	X
fuelectenerg-13681	185	3	t	t	X
fuelectenerg-13681	185	4	o	o	X
fuelectenerg-13681	185	5	t	t	NOUN
fuelectenerg-13681	185	6	o	o	NOUN
fuelectenerg-13681	185	7	−	−	PROPN
fuelectenerg-13681	185	8	−=	−=	NOUN
fuelectenerg-13681	185	9	+	+	PUNCT
fuelectenerg-13681	186	1	+	+	ADJ
fuelectenerg-13681	186	2	o	o	X
fuelectenerg-13681	186	3	w	w	NOUN
fuelectenerg-13681	186	4	h	h	NOUN
fuelectenerg-13681	186	5	u	u	NOUN
fuelectenerg-13681	186	6	h	h	PROPN
fuelectenerg-13681	186	7	b	b	PROPN
fuelectenerg-13681	186	8	(	(	PUNCT
fuelectenerg-13681	186	9	26	26	NUM
fuelectenerg-13681	186	10	)	)	PUNCT
fuelectenerg-13681	186	11	1	1	NUM
fuelectenerg-13681	186	12	1tanh	1tanh	NUM
fuelectenerg-13681	186	13	(	(	PUNCT
fuelectenerg-13681	186	14	)	)	PUNCT
fuelectenerg-13681	186	15	l	l	NOUN
fuelectenerg-13681	186	16	l	l	NOUN
fuelectenerg-13681	186	17	l	l	NOUN
fuelectenerg-13681	186	18	l	l	X
fuelectenerg-13681	186	19	l	l	X
fuelectenerg-13681	186	20	l	l	NOUN
fuelectenerg-13681	186	21	t	t	NOUN
fuelectenerg-13681	186	22	c	c	NOUN
fuelectenerg-13681	186	23	t	t	PROPN
fuelectenerg-13681	186	24	c	c	PROPN
fuelectenerg-13681	186	25	t	t	PROPN
fuelectenerg-13681	186	26	c	c	NOUN
fuelectenerg-13681	186	27	−	−	PROPN
fuelectenerg-13681	186	28	−=	−=	X
fuelectenerg-13681	186	29	+	+	PUNCT
fuelectenerg-13681	187	1	+	+	ADP
fuelectenerg-13681	187	2	c	c	PROPN
fuelectenerg-13681	187	3	w	w	NOUN
fuelectenerg-13681	187	4	h	h	NOUN
fuelectenerg-13681	187	5	u	u	NOUN
fuelectenerg-13681	187	6	h	h	PROPN
fuelectenerg-13681	187	7	b	b	PROPN
fuelectenerg-13681	187	8	(	(	PUNCT
fuelectenerg-13681	187	9	27	27	NUM
fuelectenerg-13681	187	10	)	)	PUNCT
fuelectenerg-13681	187	11	1	1	NUM
fuelectenerg-13681	187	12	l	l	NOUN
fuelectenerg-13681	187	13	l	l	NOUN
fuelectenerg-13681	187	14	l	l	X
fuelectenerg-13681	187	15	l	l	X
fuelectenerg-13681	187	16	l	l	NOUN
fuelectenerg-13681	187	17	t	t	NOUN
fuelectenerg-13681	187	18	t	t	PROPN
fuelectenerg-13681	187	19	t	t	PROPN
fuelectenerg-13681	187	20	t	t	X
fuelectenerg-13681	187	21	t−=	t−=	PROPN
fuelectenerg-13681	188	1	+	+	PROPN
fuelectenerg-13681	188	2	c	c	X
fuelectenerg-13681	188	3	f	f	NOUN
fuelectenerg-13681	189	1	c	c	NOUN
fuelectenerg-13681	189	2	i	i	PRON
fuelectenerg-13681	189	3	c	c	PROPN
fuelectenerg-13681	189	4	(	(	PUNCT
fuelectenerg-13681	189	5	28	28	NUM
fuelectenerg-13681	189	6	)	)	PUNCT
fuelectenerg-13681	189	7	tanh	tanh	NOUN
fuelectenerg-13681	189	8	(	(	PUNCT
fuelectenerg-13681	189	9	)	)	PUNCT
fuelectenerg-13681	189	10	l	l	NOUN
fuelectenerg-13681	190	1	l	l	NOUN
fuelectenerg-13681	190	2	l	l	NOUN
fuelectenerg-13681	190	3	t	t	NOUN
fuelectenerg-13681	190	4	t	t	PROPN
fuelectenerg-13681	190	5	t	t	PROPN
fuelectenerg-13681	190	6	=	=	PROPN
fuelectenerg-13681	190	7	h	h	NOUN
fuelectenerg-13681	190	8	o	o	X
fuelectenerg-13681	190	9	c	c	X
fuelectenerg-13681	190	10	(	(	PUNCT
fuelectenerg-13681	190	11	29	29	NUM
fuelectenerg-13681	190	12	)	)	PUNCT
fuelectenerg-13681	190	13	where	where	SCONJ
fuelectenerg-13681	190	14	l	l	NOUN
fuelectenerg-13681	190	15	tf	tf	X
fuelectenerg-13681	190	16	l	l	NOUN
fuelectenerg-13681	190	17	ti	ti	NOUN
fuelectenerg-13681	190	18	and	and	CCONJ
fuelectenerg-13681	190	19	l	l	NOUN
fuelectenerg-13681	190	20	to	to	PART
fuelectenerg-13681	190	21	are	are	VERB
fuelectenerg-13681	190	22	the	the	DET
fuelectenerg-13681	190	23	forget	forget	NOUN
fuelectenerg-13681	190	24	,	,	PUNCT
fuelectenerg-13681	190	25	input	input	NOUN
fuelectenerg-13681	190	26	,	,	PUNCT
fuelectenerg-13681	190	27	and	and	CCONJ
fuelectenerg-13681	190	28	output	output	NOUN
fuelectenerg-13681	190	29	gates	gate	NOUN
fuelectenerg-13681	190	30	,	,	PUNCT
fuelectenerg-13681	190	31	l	l	PROPN
fuelectenerg-13681	190	32	tc	tc	NOUN
fuelectenerg-13681	190	33	is	be	AUX
fuelectenerg-13681	190	34	the	the	DET
fuelectenerg-13681	190	35	cell	cell	NOUN
fuelectenerg-13681	190	36	state	state	NOUN
fuelectenerg-13681	190	37	,	,	PUNCT
fuelectenerg-13681	190	38	l	l	PROPN
fuelectenerg-13681	190	39	th	th	X
fuelectenerg-13681	190	40	is	be	AUX
fuelectenerg-13681	190	41	the	the	DET
fuelectenerg-13681	190	42	hidden	hidden	ADJ
fuelectenerg-13681	190	43	state	state	NOUN
fuelectenerg-13681	190	44	,	,	PUNCT
fuelectenerg-13681	190	45			PROPN
fuelectenerg-13681	190	46	is	be	AUX
fuelectenerg-13681	190	47	the	the	DET
fuelectenerg-13681	190	48	sigmoid	sigmoid	NOUN
fuelectenerg-13681	190	49	function	function	NOUN
fuelectenerg-13681	190	50	,	,	PUNCT
fuelectenerg-13681	190	51	and	and	CCONJ
fuelectenerg-13681	190	52	represents	represent	VERB
fuelectenerg-13681	190	53	element	element	ADJ
fuelectenerg-13681	190	54	-	-	ADJ
fuelectenerg-13681	190	55	wise	wise	ADJ
fuelectenerg-13681	190	56	multiplication	multiplication	NOUN
fuelectenerg-13681	190	57	.	.	PUNCT
fuelectenerg-13681	191	1	specifically	specifically	ADV
fuelectenerg-13681	191	2	,	,	PUNCT
fuelectenerg-13681	191	3	we	we	PRON
fuelectenerg-13681	191	4	unfold	unfold	VERB
fuelectenerg-13681	191	5	the	the	DET
fuelectenerg-13681	191	6	proximal	proximal	ADJ
fuelectenerg-13681	191	7	gradient	gradient	ADJ
fuelectenerg-13681	191	8	descent	descent	NOUN
fuelectenerg-13681	191	9	algorithm	algorithm	NOUN
fuelectenerg-13681	191	10	:	:	PUNCT
fuelectenerg-13681	191	11	1	1	NUM
fuelectenerg-13681	191	12	(	(	PUNCT
fuelectenerg-13681	191	13	)	)	PUNCT
fuelectenerg-13681	191	14	prox	prox	PROPN
fuelectenerg-13681	191	15	(	(	PUNCT
fuelectenerg-13681	191	16	)	)	PUNCT
fuelectenerg-13681	191	17	k	k	PROPN
fuelectenerg-13681	191	18	g	g	PROPN
fuelectenerg-13681	191	19	k	k	PROPN
fuelectenerg-13681	191	20	kf	kf	X
fuelectenerg-13681	191	21	+	+	NOUN
fuelectenerg-13681	191	22	=	=	SYM
fuelectenerg-13681	191	23	−	−	NOUN
fuelectenerg-13681	191	24	w	w	PROPN
fuelectenerg-13681	191	25	w	w	PROPN
fuelectenerg-13681	191	26	w	w	PROPN
fuelectenerg-13681	191	27	(	(	PUNCT
fuelectenerg-13681	191	28	30	30	NUM
fuelectenerg-13681	191	29	)	)	PUNCT
fuelectenerg-13681	191	30	where	where	SCONJ
fuelectenerg-13681	191	31	f	f	PROPN
fuelectenerg-13681	191	32	is	be	AUX
fuelectenerg-13681	191	33	the	the	DET
fuelectenerg-13681	191	34	data	datum	NOUN
fuelectenerg-13681	191	35	fidelity	fidelity	PROPN
fuelectenerg-13681	191	36	term	term	NOUN
fuelectenerg-13681	191	37	,	,	PUNCT
fuelectenerg-13681	191	38	g	g	PROPN
fuelectenerg-13681	191	39	is	be	AUX
fuelectenerg-13681	191	40	the	the	DET
fuelectenerg-13681	191	41	regularization	regularization	NOUN
fuelectenerg-13681	191	42	term	term	NOUN
fuelectenerg-13681	191	43	,	,	PUNCT
fuelectenerg-13681	191	44			NOUN
fuelectenerg-13681	191	45	is	be	AUX
fuelectenerg-13681	191	46	the	the	DET
fuelectenerg-13681	191	47	step	step	NOUN
fuelectenerg-13681	191	48	size	size	NOUN
fuelectenerg-13681	191	49	,	,	PUNCT
fuelectenerg-13681	191	50	and	and	CCONJ
fuelectenerg-13681	191	51	prox	prox	NOUN
fuelectenerg-13681	191	52	is	be	AUX
fuelectenerg-13681	191	53	the	the	DET
fuelectenerg-13681	191	54	proximal	proximal	ADJ
fuelectenerg-13681	191	55	operator	operator	NOUN
fuelectenerg-13681	191	56	.	.	PUNCT
fuelectenerg-13681	192	1	in	in	ADP
fuelectenerg-13681	192	2	our	our	PRON
fuelectenerg-13681	192	3	model	model	NOUN
fuelectenerg-13681	192	4	-	-	PUNCT
fuelectenerg-13681	192	5	based	base	VERB
fuelectenerg-13681	192	6	deep	deep	ADJ
fuelectenerg-13681	192	7	learning	learning	NOUN
fuelectenerg-13681	192	8	approach	approach	NOUN
fuelectenerg-13681	192	9	,	,	PUNCT
fuelectenerg-13681	192	10	we	we	PRON
fuelectenerg-13681	192	11	parameterize	parameterize	VERB
fuelectenerg-13681	192	12	this	this	PRON
fuelectenerg-13681	192	13	as	as	ADP
fuelectenerg-13681	192	14	:	:	PUNCT
fuelectenerg-13681	192	15	(	(	PUNCT
fuelectenerg-13681	192	16	)	)	SYM
fuelectenerg-13681	192	17	1	1	NUM
fuelectenerg-13681	192	18	prox	prox	NOUN
fuelectenerg-13681	192	19	(	(	PUNCT
fuelectenerg-13681	192	20	;	;	PUNCT
fuelectenerg-13681	192	21	)	)	PUNCT
fuelectenerg-13681	192	22	kk	kk	INTJ
fuelectenerg-13681	192	23	g	g	PROPN
fuelectenerg-13681	192	24	k	k	PROPN
fuelectenerg-13681	192	25	k	k	PROPN
fuelectenerg-13681	192	26	k	k	PROPN
fuelectenerg-13681	192	27	kf	kf	X
fuelectenerg-13681	192	28			X
fuelectenerg-13681	192	29	+	+	VERB
fuelectenerg-13681	192	30	=	=	SYM
fuelectenerg-13681	193	1	−	−	NOUN
fuelectenerg-13681	193	2	w	w	PROPN
fuelectenerg-13681	193	3	w	w	PROPN
fuelectenerg-13681	193	4	w	w	PROPN
fuelectenerg-13681	193	5	(	(	PUNCT
fuelectenerg-13681	193	6	31	31	NUM
fuelectenerg-13681	193	7	)	)	PUNCT
fuelectenerg-13681	193	8	where	where	SCONJ
fuelectenerg-13681	193	9	k	k	NUM
fuelectenerg-13681	193	10	and	and	CCONJ
fuelectenerg-13681	193	11	k	k	VERB
fuelectenerg-13681	193	12	are	be	AUX
fuelectenerg-13681	193	13	learnable	learnable	ADJ
fuelectenerg-13681	193	14	parameters	parameter	NOUN
fuelectenerg-13681	193	15	for	for	ADP
fuelectenerg-13681	193	16	the	the	DET
fuelectenerg-13681	193	17	k	k	PROPN
fuelectenerg-13681	193	18	h	h	PROPN
fuelectenerg-13681	193	19	iteration	iteration	NOUN
fuelectenerg-13681	193	20	/	/	SYM
fuelectenerg-13681	193	21	layer	layer	NOUN
fuelectenerg-13681	193	22	.	.	PUNCT
fuelectenerg-13681	194	1	all	all	DET
fuelectenerg-13681	194	2	networks	network	NOUN
fuelectenerg-13681	194	3	are	be	AUX
fuelectenerg-13681	194	4	implemented	implement	VERB
fuelectenerg-13681	194	5	in	in	ADP
fuelectenerg-13681	194	6	tensorflow	tensorflow	NOUN
fuelectenerg-13681	194	7	2.5	2.5	NUM
fuelectenerg-13681	194	8	and	and	CCONJ
fuelectenerg-13681	194	9	trained	train	VERB
fuelectenerg-13681	194	10	using	use	VERB
fuelectenerg-13681	194	11	the	the	DET
fuelectenerg-13681	194	12	adam	adam	PROPN
fuelectenerg-13681	194	13	optimizer	optimizer	NOUN
fuelectenerg-13681	194	14	with	with	ADP
fuelectenerg-13681	194	15	a	a	DET
fuelectenerg-13681	194	16	learning	learn	VERB
fuelectenerg-13681	194	17	rate	rate	NOUN
fuelectenerg-13681	194	18	of	of	ADP
fuelectenerg-13681	194	19	0.001	0.001	NUM
fuelectenerg-13681	194	20	,	,	PUNCT
fuelectenerg-13681	194	21	batch	batch	NOUN
fuelectenerg-13681	194	22	size	size	NOUN
fuelectenerg-13681	194	23	of	of	ADP
fuelectenerg-13681	194	24	128	128	NUM
fuelectenerg-13681	194	25	,	,	PUNCT
fuelectenerg-13681	194	26	and	and	CCONJ
fuelectenerg-13681	194	27	early	early	ADV
fuelectenerg-13681	194	28	stopping	stop	VERB
fuelectenerg-13681	194	29	with	with	ADP
fuelectenerg-13681	194	30	a	a	DET
fuelectenerg-13681	194	31	patience	patience	NOUN
fuelectenerg-13681	194	32	of	of	ADP
fuelectenerg-13681	194	33	10	10	NUM
fuelectenerg-13681	194	34	epochs	epoch	NOUN
fuelectenerg-13681	194	35	.	.	PUNCT
fuelectenerg-13681	195	1	to	to	PART
fuelectenerg-13681	195	2	ensure	ensure	VERB
fuelectenerg-13681	195	3	reproducibility	reproducibility	NOUN
fuelectenerg-13681	195	4	,	,	PUNCT
fuelectenerg-13681	195	5	we	we	PRON
fuelectenerg-13681	195	6	use	use	VERB
fuelectenerg-13681	195	7	fixed	fix	VERB
fuelectenerg-13681	195	8	random	random	ADJ
fuelectenerg-13681	195	9	seeds	seed	NOUN
fuelectenerg-13681	195	10	and	and	CCONJ
fuelectenerg-13681	195	11	save	save	VERB
fuelectenerg-13681	195	12	model	model	NOUN
fuelectenerg-13681	195	13	checkpoints	checkpoint	NOUN
fuelectenerg-13681	195	14	at	at	ADP
fuelectenerg-13681	195	15	regular	regular	ADJ
fuelectenerg-13681	195	16	intervals	interval	NOUN
fuelectenerg-13681	195	17	.	.	PUNCT
fuelectenerg-13681	196	1	3.4	3.4	NUM
fuelectenerg-13681	196	2	.	.	PUNCT
fuelectenerg-13681	196	3	hybrid	hybrid	ADJ
fuelectenerg-13681	196	4	approach	approach	NOUN
fuelectenerg-13681	196	5	we	we	PRON
fuelectenerg-13681	196	6	propose	propose	VERB
fuelectenerg-13681	196	7	a	a	DET
fuelectenerg-13681	196	8	novel	novel	ADJ
fuelectenerg-13681	196	9	hybrid	hybrid	NOUN
fuelectenerg-13681	196	10	approach	approach	NOUN
fuelectenerg-13681	196	11	that	that	PRON
fuelectenerg-13681	196	12	dynamically	dynamically	ADV
fuelectenerg-13681	196	13	selects	select	NOUN
fuelectenerg-13681	196	14	between	between	ADP
fuelectenerg-13681	196	15	traditional	traditional	ADJ
fuelectenerg-13681	196	16	and	and	CCONJ
fuelectenerg-13681	196	17	deep	deep	ADJ
fuelectenerg-13681	196	18	learning	learning	NOUN
fuelectenerg-13681	196	19	methods	method	NOUN
fuelectenerg-13681	196	20	based	base	VERB
fuelectenerg-13681	196	21	on	on	ADP
fuelectenerg-13681	196	22	channel	channel	NOUN
fuelectenerg-13681	196	23	conditions	condition	NOUN
fuelectenerg-13681	196	24	,	,	PUNCT
fuelectenerg-13681	196	25	mobility	mobility	NOUN
fuelectenerg-13681	196	26	patterns	pattern	NOUN
fuelectenerg-13681	196	27	,	,	PUNCT
fuelectenerg-13681	196	28	and	and	CCONJ
fuelectenerg-13681	196	29	computational	computational	ADJ
fuelectenerg-13681	196	30	constraints	constraint	NOUN
fuelectenerg-13681	196	31	.	.	PUNCT
fuelectenerg-13681	197	1	the	the	DET
fuelectenerg-13681	197	2	key	key	ADJ
fuelectenerg-13681	197	3	innovation	innovation	NOUN
fuelectenerg-13681	197	4	in	in	ADP
fuelectenerg-13681	197	5	our	our	PRON
fuelectenerg-13681	197	6	approach	approach	NOUN
fuelectenerg-13681	197	7	is	be	AUX
fuelectenerg-13681	197	8	the	the	DET
fuelectenerg-13681	197	9	formulation	formulation	NOUN
fuelectenerg-13681	197	10	of	of	ADP
fuelectenerg-13681	197	11	the	the	DET
fuelectenerg-13681	197	12	algorithm	algorithm	NOUN
fuelectenerg-13681	197	13	selection	selection	NOUN
fuelectenerg-13681	197	14	problem	problem	NOUN
fuelectenerg-13681	197	15	as	as	ADP
fuelectenerg-13681	197	16	a	a	DET
fuelectenerg-13681	197	17	multi	multi	ADJ
fuelectenerg-13681	197	18	-	-	ADJ
fuelectenerg-13681	197	19	armed	armed	ADJ
fuelectenerg-13681	197	20	bandit	bandit	NOUN
fuelectenerg-13681	197	21	problem	problem	NOUN
fuelectenerg-13681	197	22	,	,	PUNCT
fuelectenerg-13681	197	23	which	which	PRON
fuelectenerg-13681	197	24	allows	allow	VERB
fuelectenerg-13681	197	25	for	for	ADP
fuelectenerg-13681	197	26	online	online	ADJ
fuelectenerg-13681	197	27	learning	learning	NOUN
fuelectenerg-13681	197	28	of	of	ADP
fuelectenerg-13681	197	29	the	the	DET
fuelectenerg-13681	197	30	optimal	optimal	ADJ
fuelectenerg-13681	197	31	strategy	strategy	NOUN
fuelectenerg-13681	197	32	while	while	SCONJ
fuelectenerg-13681	197	33	balancing	balance	VERB
fuelectenerg-13681	197	34	exploration	exploration	NOUN
fuelectenerg-13681	197	35	and	and	CCONJ
fuelectenerg-13681	197	36	exploitation	exploitation	NOUN
fuelectenerg-13681	197	37	.	.	PUNCT
fuelectenerg-13681	198	1	formally	formally	ADV
fuelectenerg-13681	198	2	,	,	PUNCT
fuelectenerg-13681	198	3	let	let	VERB
fuelectenerg-13681	198	4	a	a	DET
fuelectenerg-13681	198	5	=	=	PUNCT
fuelectenerg-13681	198	6	{	{	PUNCT
fuelectenerg-13681	198	7	a1	a1	PROPN
fuelectenerg-13681	198	8	,	,	PUNCT
fuelectenerg-13681	198	9	a2,	a2,	ADP
fuelectenerg-13681	198	10	…	…	X
fuelectenerg-13681	198	11	,ak	,ak	PUNCT
fuelectenerg-13681	198	12	}	}	PUNCT
fuelectenerg-13681	198	13	be	be	VERB
fuelectenerg-13681	198	14	the	the	DET
fuelectenerg-13681	198	15	set	set	NOUN
fuelectenerg-13681	198	16	of	of	ADP
fuelectenerg-13681	198	17	available	available	ADJ
fuelectenerg-13681	198	18	algorithms	algorithm	NOUN
fuelectenerg-13681	198	19	,	,	PUNCT
fuelectenerg-13681	198	20	where	where	SCONJ
fuelectenerg-13681	198	21	each	each	DET
fuelectenerg-13681	198	22	algorithm	algorithm	NOUN
fuelectenerg-13681	198	23	ai	ai	VERB
fuelectenerg-13681	198	24	can	can	AUX
fuelectenerg-13681	198	25	be	be	AUX
fuelectenerg-13681	198	26	either	either	CCONJ
fuelectenerg-13681	198	27	a	a	DET
fuelectenerg-13681	198	28	traditional	traditional	ADJ
fuelectenerg-13681	198	29	optimization	optimization	NOUN
fuelectenerg-13681	198	30	method	method	NOUN
fuelectenerg-13681	198	31	or	or	CCONJ
fuelectenerg-13681	198	32	a	a	DET
fuelectenerg-13681	198	33	deep	deep	ADJ
fuelectenerg-13681	198	34	learning	learning	NOUN
fuelectenerg-13681	198	35	approach	approach	NOUN
fuelectenerg-13681	198	36	.	.	PUNCT
fuelectenerg-13681	199	1	for	for	ADP
fuelectenerg-13681	199	2	a	a	DET
fuelectenerg-13681	199	3	given	give	VERB
fuelectenerg-13681	199	4	channel	channel	NOUN
fuelectenerg-13681	199	5	condition	condition	NOUN
fuelectenerg-13681	199	6	and	and	CCONJ
fuelectenerg-13681	199	7	system	system	NOUN
fuelectenerg-13681	199	8	state	state	PROPN
fuelectenerg-13681	199	9	st	st	PROPN
fuelectenerg-13681	199	10	at	at	ADP
fuelectenerg-13681	199	11	time	time	NOUN
fuelectenerg-13681	199	12	t	t	PROPN
fuelectenerg-13681	199	13	the	the	DET
fuelectenerg-13681	199	14	hybrid	hybrid	ADJ
fuelectenerg-13681	199	15	approach	approach	NOUN
fuelectenerg-13681	199	16	selects	select	VERB
fuelectenerg-13681	199	17	an	an	DET
fuelectenerg-13681	199	18	algorithm	algorithm	NOUN
fuelectenerg-13681	199	19	at	at	ADP
fuelectenerg-13681	199	20			PROPN
fuelectenerg-13681	199	21	a	a	PRON
fuelectenerg-13681	199	22	to	to	PART
fuelectenerg-13681	199	23	maximize	maximize	VERB
fuelectenerg-13681	199	24	the	the	DET
fuelectenerg-13681	199	25	expected	expect	VERB
fuelectenerg-13681	199	26	performance	performance	NOUN
fuelectenerg-13681	199	27	metric	metric	ADJ
fuelectenerg-13681	199	28	rt	rt	PROPN
fuelectenerg-13681	199	29	e.g.	e.g.	ADV
fuelectenerg-13681	199	30	,	,	PUNCT
fuelectenerg-13681	199	31	sinr	sinr	NOUN
fuelectenerg-13681	199	32	)	)	PUNCT
fuelectenerg-13681	199	33	.	.	PUNCT
fuelectenerg-13681	200	1	we	we	PRON
fuelectenerg-13681	200	2	implement	implement	VERB
fuelectenerg-13681	200	3	the	the	DET
fuelectenerg-13681	200	4	selection	selection	NOUN
fuelectenerg-13681	200	5	mechanism	mechanism	NOUN
fuelectenerg-13681	200	6	using	use	VERB
fuelectenerg-13681	200	7	a	a	DET
fuelectenerg-13681	200	8	contextual	contextual	ADJ
fuelectenerg-13681	200	9	bandits	bandit	NOUN
fuelectenerg-13681	200	10	framework	framework	NOUN
fuelectenerg-13681	200	11	with	with	ADP
fuelectenerg-13681	200	12	thompson	thompson	NOUN
fuelectenerg-13681	200	13	sampling	sampling	NOUN
fuelectenerg-13681	200	14	.	.	PUNCT
fuelectenerg-13681	201	1	for	for	ADP
fuelectenerg-13681	201	2	each	each	DET
fuelectenerg-13681	201	3	algorithm	algorithm	NOUN
fuelectenerg-13681	201	4	ai	ai	VERB
fuelectenerg-13681	201	5	we	we	PRON
fuelectenerg-13681	201	6	maintain	maintain	VERB
fuelectenerg-13681	201	7	a	a	DET
fuelectenerg-13681	201	8	probabilistic	probabilistic	ADJ
fuelectenerg-13681	201	9	model	model	NOUN
fuelectenerg-13681	201	10	of	of	ADP
fuelectenerg-13681	201	11	its	its	PRON
fuelectenerg-13681	201	12	performance	performance	NOUN
fuelectenerg-13681	201	13	conditioned	condition	VERB
fuelectenerg-13681	201	14	on	on	ADP
fuelectenerg-13681	201	15	the	the	DET
fuelectenerg-13681	201	16	context	context	NOUN
fuelectenerg-13681	201	17	(	(	PUNCT
fuelectenerg-13681	201	18	channel	channel	NOUN
fuelectenerg-13681	201	19	and	and	CCONJ
fuelectenerg-13681	201	20	system	system	NOUN
fuelectenerg-13681	201	21	state	state	NOUN
fuelectenerg-13681	201	22	)	)	PUNCT
fuelectenerg-13681	201	23	.	.	PUNCT
fuelectenerg-13681	202	1	the	the	DET
fuelectenerg-13681	202	2	selection	selection	NOUN
fuelectenerg-13681	202	3	probability	probability	NOUN
fuelectenerg-13681	202	4	is	be	AUX
fuelectenerg-13681	202	5	given	give	VERB
fuelectenerg-13681	202	6	by	by	ADP
fuelectenerg-13681	202	7	:	:	PUNCT
fuelectenerg-13681	202	8	)	)	PUNCT
fuelectenerg-13681	202	9	(	(	PUNCT
fuelectenerg-13681	202	10	)	)	PUNCT
fuelectenerg-13681	202	11	(	(	PUNCT
fuelectenerg-13681	202	12	|	|	ADV
fuelectenerg-13681	202	13	)	)	PUNCT
fuelectenerg-13681	202	14	(	(	PUNCT
fuelectenerg-13681	202	15	,	,	PUNCT
fuelectenerg-13681	202	16	(	(	PUNCT
fuelectenerg-13681	202	17	)	)	PUNCT
fuelectenerg-13681	202	18	t	t	NOUN
fuelectenerg-13681	203	1	i	i	PRON
fuelectenerg-13681	203	2	t	t	X
fuelectenerg-13681	204	1	i	i	PRON
fuelectenerg-13681	204	2	t	t	PROPN
fuelectenerg-13681	204	3	j	j	PROPN
fuelectenerg-13681	204	4	tp	tp	ADP
fuelectenerg-13681	204	5	a	a	DET
fuelectenerg-13681	204	6	a	a	DET
fuelectenerg-13681	204	7	s	s	X
fuelectenerg-13681	204	8	p	p	NOUN
fuelectenerg-13681	204	9	r	r	NOUN
fuelectenerg-13681	204	10	s	s	NOUN
fuelectenerg-13681	204	11	r	r	NOUN
fuelectenerg-13681	204	12	s	s	PROPN
fuelectenerg-13681	204	13	j	j	NOUN
fuelectenerg-13681	204	14	i=	i=	PROPN
fuelectenerg-13681	204	15	=	=	PUNCT
fuelectenerg-13681	205	1			PROPN
fuelectenerg-13681	205	2			PROPN
fuelectenerg-13681	205	3			PROPN
fuelectenerg-13681	205	4	(	(	PUNCT
fuelectenerg-13681	205	5	32	32	NUM
fuelectenerg-13681	205	6	)	)	PUNCT
fuelectenerg-13681	205	7	where	where	SCONJ
fuelectenerg-13681	205	8	ri(st	ri(st	NOUN
fuelectenerg-13681	205	9	)	)	PUNCT
fuelectenerg-13681	205	10	is	be	AUX
fuelectenerg-13681	205	11	the	the	DET
fuelectenerg-13681	205	12	reward	reward	NOUN
fuelectenerg-13681	205	13	(	(	PUNCT
fuelectenerg-13681	205	14	performance	performance	NOUN
fuelectenerg-13681	205	15	)	)	PUNCT
fuelectenerg-13681	205	16	of	of	ADP
fuelectenerg-13681	205	17	algorithm	algorithm	NOUN
fuelectenerg-13681	205	18	ai	ai	VERB
fuelectenerg-13681	205	19	in	in	ADP
fuelectenerg-13681	205	20	state	state	PROPN
fuelectenerg-13681	205	21	st	st	PROPN
fuelectenerg-13681	205	22	.	.	PROPN
fuelectenerg-13681	206	1	the	the	DET
fuelectenerg-13681	206	2	proposed	propose	VERB
fuelectenerg-13681	206	3	hybrid	hybrid	ADJ
fuelectenerg-13681	206	4	algorithm	algorithm	NOUN
fuelectenerg-13681	206	5	selection	selection	NOUN
fuelectenerg-13681	206	6	approach	approach	NOUN
fuelectenerg-13681	206	7	is	be	AUX
fuelectenerg-13681	206	8	detailed	detail	VERB
fuelectenerg-13681	206	9	in	in	ADP
fuelectenerg-13681	206	10	table	table	NOUN
fuelectenerg-13681	206	11	1	1	NUM
fuelectenerg-13681	206	12	,	,	PUNCT
fuelectenerg-13681	206	13	which	which	PRON
fuelectenerg-13681	206	14	outlines	outline	VERB
fuelectenerg-13681	206	15	the	the	DET
fuelectenerg-13681	206	16	process	process	NOUN
fuelectenerg-13681	206	17	for	for	ADP
fuelectenerg-13681	206	18	dynamically	dynamically	ADV
fuelectenerg-13681	206	19	selecting	select	VERB
fuelectenerg-13681	206	20	the	the	DET
fuelectenerg-13681	206	21	most	most	ADV
fuelectenerg-13681	206	22	appropriate	appropriate	ADJ
fuelectenerg-13681	206	23	beamforming	beamforme	VERB
fuelectenerg-13681	206	24	algorithm	algorithm	NOUN
fuelectenerg-13681	206	25	based	base	VERB
fuelectenerg-13681	206	26	on	on	ADP
fuelectenerg-13681	206	27	current	current	ADJ
fuelectenerg-13681	206	28	channel	channel	NOUN
fuelectenerg-13681	206	29	conditions	condition	NOUN
fuelectenerg-13681	206	30	.	.	PUNCT
fuelectenerg-13681	207	1	206	206	NUM
fuelectenerg-13681	207	2	m.	m.	NOUN
fuelectenerg-13681	207	3	mohamed	mohamed	PROPN
fuelectenerg-13681	207	4	,	,	PUNCT
fuelectenerg-13681	207	5	f.	f.	PROPN
fuelectenerg-13681	207	6	aljuaid	aljuaid	PROPN
fuelectenerg-13681	207	7	,	,	PUNCT
fuelectenerg-13681	207	8	mhd	mhd	PROPN
fuelectenerg-13681	207	9	w.	w.	PROPN
fuelectenerg-13681	207	10	koubeisi	koubeisi	PROPN
fuelectenerg-13681	207	11	table	table	NOUN
fuelectenerg-13681	207	12	1	1	NUM
fuelectenerg-13681	207	13	hybrid	hybrid	ADJ
fuelectenerg-13681	207	14	algorithm	algorithm	NOUN
fuelectenerg-13681	207	15	selection	selection	NOUN
fuelectenerg-13681	207	16	algorithm	algorithm	NOUN
fuelectenerg-13681	207	17	1	1	NUM
fuelectenerg-13681	207	18	hybrid	hybrid	ADJ
fuelectenerg-13681	207	19	algorithm	algorithm	NOUN
fuelectenerg-13681	207	20	selection	selection	NOUN
fuelectenerg-13681	207	21	input	input	NOUN
fuelectenerg-13681	207	22	:	:	PUNCT
fuelectenerg-13681	207	23	channel	channel	PROPN
fuelectenerg-13681	207	24	state	state	PROPN
fuelectenerg-13681	207	25	ht	ht	PROPN
fuelectenerg-13681	207	26	,	,	PUNCT
fuelectenerg-13681	207	27	system	system	NOUN
fuelectenerg-13681	207	28	state	state	PROPN
fuelectenerg-13681	207	29	st	st	PROPN
fuelectenerg-13681	207	30	output	output	PROPN
fuelectenerg-13681	207	31	:	:	PUNCT
fuelectenerg-13681	207	32	selected	select	VERB
fuelectenerg-13681	207	33	algorithm	algorithm	NOUN
fuelectenerg-13681	207	34	at	at	ADP
fuelectenerg-13681	207	35	and	and	CCONJ
fuelectenerg-13681	207	36	beamforming	beamforme	VERB
fuelectenerg-13681	207	37	matrix	matrix	NOUN
fuelectenerg-13681	207	38	wt	wt	ADP
fuelectenerg-13681	207	39	1	1	NUM
fuelectenerg-13681	207	40	:	:	PUNCT
fuelectenerg-13681	207	41	feature	feature	NOUN
fuelectenerg-13681	207	42	extraction	extraction	NOUN
fuelectenerg-13681	207	43	:	:	PUNCT
fuelectenerg-13681	207	44	2	2	NUM
fuelectenerg-13681	207	45	:	:	PUNCT
fuelectenerg-13681	207	46	extract	extract	VERB
fuelectenerg-13681	207	47	feature	feature	NOUN
fuelectenerg-13681	207	48	vector	vector	NOUN
fuelectenerg-13681	207	49	ft	ft	NOUN
fuelectenerg-13681	207	50	from	from	ADP
fuelectenerg-13681	207	51	ht	ht	PROPN
fuelectenerg-13681	207	52	and	and	CCONJ
fuelectenerg-13681	207	53	st	st	PROPN
fuelectenerg-13681	207	54	3	3	NUM
fuelectenerg-13681	207	55	:	:	PUNCT
fuelectenerg-13681	207	56	compute	compute	NOUN
fuelectenerg-13681	207	57	mobility	mobility	NOUN
fuelectenerg-13681	207	58	metric	metric	NOUN
fuelectenerg-13681	207	59	:	:	PUNCT
fuelectenerg-13681	208	1	mt	mt	PROPN
fuelectenerg-13681	208	2	=	=	PUNCT
fuelectenerg-13681	208	3	||ht	||ht	ADJ
fuelectenerg-13681	208	4	−ht−1||f	−ht−1||f	PROPN
fuelectenerg-13681	208	5	/	/	SYM
fuelectenerg-13681	208	6	||ht−1||f	||ht−1||f	NUM
fuelectenerg-13681	208	7	4	4	NUM
fuelectenerg-13681	208	8	:	:	PUNCT
fuelectenerg-13681	208	9	compute	compute	ADJ
fuelectenerg-13681	208	10	channel	channel	NOUN
fuelectenerg-13681	208	11	condition	condition	NOUN
fuelectenerg-13681	208	12	number	number	NOUN
fuelectenerg-13681	208	13	:	:	PUNCT
fuelectenerg-13681	208	14	t	t	PROPN
fuelectenerg-13681	208	15	=	=	SYM
fuelectenerg-13681	208	16	max(ht	max(ht	NOUN
fuelectenerg-13681	208	17	)	)	PUNCT
fuelectenerg-13681	208	18	/	/	SYM
fuelectenerg-13681	208	19	min(ht	min(ht	NOUN
fuelectenerg-13681	208	20	)	)	PUNCT
fuelectenerg-13681	208	21	5	5	NUM
fuelectenerg-13681	208	22	:	:	PUNCT
fuelectenerg-13681	208	23	compute	compute	NOUN
fuelectenerg-13681	208	24	interference	interference	NOUN
fuelectenerg-13681	208	25	-	-	PUNCT
fuelectenerg-13681	208	26	to	to	ADP
fuelectenerg-13681	208	27	-	-	PUNCT
fuelectenerg-13681	208	28	signal	signal	NOUN
fuelectenerg-13681	208	29	ratio	ratio	NOUN
fuelectenerg-13681	208	30	:	:	PUNCT
fuelectenerg-13681	208	31	t	t	NOUN
fuelectenerg-13681	208	32	=	=	SYM
fuelectenerg-13681	208	33	tr(ri	tr(ri	PROPN
fuelectenerg-13681	208	34	)	)	PUNCT
fuelectenerg-13681	208	35	/	/	SYM
fuelectenerg-13681	208	36	tr(rs	tr(rs	NOUN
fuelectenerg-13681	208	37	)	)	PUNCT
fuelectenerg-13681	208	38	6	6	NUM
fuelectenerg-13681	208	39	:	:	PUNCT
fuelectenerg-13681	208	40	estimate	estimate	VERB
fuelectenerg-13681	208	41	computational	computational	ADJ
fuelectenerg-13681	208	42	budget	budget	NOUN
fuelectenerg-13681	208	43	bt	bt	NOUN
fuelectenerg-13681	208	44	for	for	ADP
fuelectenerg-13681	208	45	current	current	ADJ
fuelectenerg-13681	208	46	time	time	NOUN
fuelectenerg-13681	208	47	slot	slot	NOUN
fuelectenerg-13681	208	48	7	7	NUM
fuelectenerg-13681	208	49	:	:	PUNCT
fuelectenerg-13681	208	50	algorithm	algorithm	NOUN
fuelectenerg-13681	208	51	selection	selection	NOUN
fuelectenerg-13681	208	52	using	use	VERB
fuelectenerg-13681	208	53	thompson	thompson	NOUN
fuelectenerg-13681	208	54	sampling	sampling	NOUN
fuelectenerg-13681	208	55	:	:	PUNCT
fuelectenerg-13681	208	56	8	8	NUM
fuelectenerg-13681	208	57	:	:	PUNCT
fuelectenerg-13681	208	58	for	for	ADP
fuelectenerg-13681	208	59	each	each	DET
fuelectenerg-13681	208	60	algorithm	algorithm	NOUN
fuelectenerg-13681	208	61	ai	ai	VERB
fuelectenerg-13681	208	62			PROPN
fuelectenerg-13681	208	63	a	a	DET
fuelectenerg-13681	208	64	do	do	PROPN
fuelectenerg-13681	208	65	9	9	NUM
fuelectenerg-13681	208	66	:	:	PUNCT
fuelectenerg-13681	208	67	sample	sample	NOUN
fuelectenerg-13681	208	68	expected	expect	VERB
fuelectenerg-13681	208	69	reward	reward	NOUN
fuelectenerg-13681	208	70	:	:	PUNCT
fuelectenerg-13681	208	71	)	)	PUNCT
fuelectenerg-13681	208	72	)	)	PUNCT
fuelectenerg-13681	208	73	(	(	PUNCT
fuelectenerg-13681	208	74	)	)	PUNCT
fuelectenerg-13681	208	75	,	,	PUNCT
fuelectenerg-13681	208	76	(	(	PUNCT
fuelectenerg-13681	208	77	(	(	PUNCT
fuelectenerg-13681	208	78	~~	~~	SYM
fuelectenerg-13681	208	79	2	2	NUM
fuelectenerg-13681	208	80	titii	titii	NOUN
fuelectenerg-13681	208	81	ffnr	ffnr	NOUN
fuelectenerg-13681	208	82			PROPN
fuelectenerg-13681	208	83	10	10	NUM
fuelectenerg-13681	208	84	:	:	PUNCT
fuelectenerg-13681	208	85	estimate	estimate	VERB
fuelectenerg-13681	208	86	execution	execution	NOUN
fuelectenerg-13681	208	87	time	time	NOUN
fuelectenerg-13681	208	88	:	:	PUNCT
fuelectenerg-13681	208	89	τi	τi	ADP
fuelectenerg-13681	208	90	=	=	SYM
fuelectenerg-13681	208	91	executiontimemodel(αi	executiontimemodel(αi	ADV
fuelectenerg-13681	208	92	,	,	PUNCT
fuelectenerg-13681	208	93	ft	ft	NOUN
fuelectenerg-13681	208	94	)	)	PUNCT
fuelectenerg-13681	208	95	11	11	NUM
fuelectenerg-13681	208	96	:	:	PUNCT
fuelectenerg-13681	208	97	if	if	SCONJ
fuelectenerg-13681	208	98	i	i	NOUN
fuelectenerg-13681	208	99	>	>	X
fuelectenerg-13681	208	100	bt	bt	NOUN
fuelectenerg-13681	208	101	then	then	ADV
fuelectenerg-13681	208	102	12	12	NUM
fuelectenerg-13681	208	103	:	:	PUNCT
fuelectenerg-13681	208	104	−=ir	−=ir	NOUN
fuelectenerg-13681	208	105	~	~	PUNCT
fuelectenerg-13681	208	106	//	//	X
fuelectenerg-13681	208	107	exclude	exclude	VERB
fuelectenerg-13681	208	108	algorithms	algorithm	NOUN
fuelectenerg-13681	208	109	exceeding	exceed	VERB
fuelectenerg-13681	208	110	budget	budget	NOUN
fuelectenerg-13681	208	111	13	13	NUM
fuelectenerg-13681	208	112	:	:	PUNCT
fuelectenerg-13681	208	113	end	end	VERB
fuelectenerg-13681	208	114	if	if	SCONJ
fuelectenerg-13681	208	115	14	14	NUM
fuelectenerg-13681	208	116	:	:	PUNCT
fuelectenerg-13681	208	117	end	end	VERB
fuelectenerg-13681	208	118	for	for	ADP
fuelectenerg-13681	208	119	15	15	NUM
fuelectenerg-13681	208	120	:	:	PUNCT
fuelectenerg-13681	208	121	execute	execute	VERB
fuelectenerg-13681	208	122	selected	select	VERB
fuelectenerg-13681	208	123	algorithm	algorithm	NOUN
fuelectenerg-13681	208	124	and	and	CCONJ
fuelectenerg-13681	208	125	update	update	NOUN
fuelectenerg-13681	208	126	model	model	NOUN
fuelectenerg-13681	208	127	:	:	PUNCT
fuelectenerg-13681	208	128	16	16	NUM
fuelectenerg-13681	208	129	:	:	PUNCT
fuelectenerg-13681	208	130	select	select	ADJ
fuelectenerg-13681	208	131	algorithm	algorithm	NOUN
fuelectenerg-13681	208	132	:	:	PUNCT
fuelectenerg-13681	209	1	iaat	iaat	PROPN
fuelectenerg-13681	209	2	ra	ra	PROPN
fuelectenerg-13681	210	1	i	i	PRON
fuelectenerg-13681	210	2	~maxarg	~maxarg	PUNCT
fuelectenerg-13681	210	3	=	=	VERB
fuelectenerg-13681	210	4	17	17	NUM
fuelectenerg-13681	210	5	:	:	PUNCT
fuelectenerg-13681	210	6	execute	execute	VERB
fuelectenerg-13681	210	7	algorithm	algorithm	NOUN
fuelectenerg-13681	210	8	at	at	ADP
fuelectenerg-13681	210	9	to	to	PART
fuelectenerg-13681	210	10	obtain	obtain	VERB
fuelectenerg-13681	210	11	beamforming	beamforming	NOUN
fuelectenerg-13681	210	12	matrix	matrix	NOUN
fuelectenerg-13681	210	13	wt	wt	ADP
fuelectenerg-13681	210	14	18	18	NUM
fuelectenerg-13681	210	15	:	:	PUNCT
fuelectenerg-13681	210	16	observe	observe	VERB
fuelectenerg-13681	210	17	actual	actual	ADJ
fuelectenerg-13681	210	18	performance	performance	NOUN
fuelectenerg-13681	210	19	rt	rt	PROPN
fuelectenerg-13681	210	20	19	19	NUM
fuelectenerg-13681	210	21	:	:	PUNCT
fuelectenerg-13681	210	22	update	update	VERB
fuelectenerg-13681	210	23	probabilistic	probabilistic	ADJ
fuelectenerg-13681	210	24	model	model	NOUN
fuelectenerg-13681	210	25	parameters	parameter	NOUN
fuelectenerg-13681	210	26	for	for	ADP
fuelectenerg-13681	210	27	algorithm	algorithm	NOUN
fuelectenerg-13681	210	28	at	at	ADP
fuelectenerg-13681	210	29	:	:	PUNCT
fuelectenerg-13681	210	30	20	20	NUM
fuelectenerg-13681	210	31	:	:	PUNCT
fuelectenerg-13681	210	32	µi(ft	µi(ft	ADJ
fuelectenerg-13681	210	33	)	)	PUNCT
fuelectenerg-13681	210	34	=	=	SYM
fuelectenerg-13681	211	1	updatemean(µi(ft	updatemean(µi(ft	ADJ
fuelectenerg-13681	211	2	)	)	PUNCT
fuelectenerg-13681	211	3	,	,	PUNCT
fuelectenerg-13681	211	4	rt	rt	PROPN
fuelectenerg-13681	211	5	,	,	PUNCT
fuelectenerg-13681	211	6	ft	ft	PROPN
fuelectenerg-13681	211	7	)	)	PUNCT
fuelectenerg-13681	211	8	21	21	NUM
fuelectenerg-13681	211	9	:	:	PUNCT
fuelectenerg-13681	211	10	σi	σi	NUM
fuelectenerg-13681	211	11	2(ft	2(ft	NUM
fuelectenerg-13681	211	12	)	)	PUNCT
fuelectenerg-13681	211	13	=	=	PUNCT
fuelectenerg-13681	212	1	updatevariance(σi	updatevariance(σi	NOUN
fuelectenerg-13681	212	2	2(ft	2(ft	NUM
fuelectenerg-13681	212	3	)	)	PUNCT
fuelectenerg-13681	212	4	,	,	PUNCT
fuelectenerg-13681	212	5	rt	rt	PROPN
fuelectenerg-13681	212	6	,	,	PUNCT
fuelectenerg-13681	212	7	ft	ft	PROPN
fuelectenerg-13681	212	8	)	)	PUNCT
fuelectenerg-13681	212	9	22	22	NUM
fuelectenerg-13681	212	10	:	:	PUNCT
fuelectenerg-13681	212	11	return	return	VERB
fuelectenerg-13681	212	12	at	at	ADP
fuelectenerg-13681	212	13	,	,	PUNCT
fuelectenerg-13681	212	14	wt	wt	ADP
fuelectenerg-13681	212	15	the	the	DET
fuelectenerg-13681	212	16	selection	selection	NOUN
fuelectenerg-13681	212	17	mechanism	mechanism	NOUN
fuelectenerg-13681	212	18	uses	use	VERB
fuelectenerg-13681	212	19	several	several	ADJ
fuelectenerg-13681	212	20	key	key	ADJ
fuelectenerg-13681	212	21	features	feature	NOUN
fuelectenerg-13681	212	22	including	include	VERB
fuelectenerg-13681	212	23	mobility	mobility	NOUN
fuelectenerg-13681	212	24	metric	metric	NOUN
fuelectenerg-13681	212	25	that	that	PRON
fuelectenerg-13681	212	26	quantifies	quantify	VERB
fuelectenerg-13681	212	27	the	the	DET
fuelectenerg-13681	212	28	channel	channel	NOUN
fuelectenerg-13681	212	29	variation	variation	NOUN
fuelectenerg-13681	212	30	rate	rate	NOUN
fuelectenerg-13681	212	31	,	,	PUNCT
fuelectenerg-13681	212	32	favoring	favor	VERB
fuelectenerg-13681	212	33	deep	deep	ADJ
fuelectenerg-13681	212	34	learning	learning	NOUN
fuelectenerg-13681	212	35	in	in	ADP
fuelectenerg-13681	212	36	high	high	ADJ
fuelectenerg-13681	212	37	-	-	PUNCT
fuelectenerg-13681	212	38	mobility	mobility	NOUN
fuelectenerg-13681	212	39	scenarios	scenario	NOUN
fuelectenerg-13681	212	40	;	;	PUNCT
fuelectenerg-13681	212	41	channel	channel	NOUN
fuelectenerg-13681	212	42	condition	condition	NOUN
fuelectenerg-13681	212	43	number	number	NOUN
fuelectenerg-13681	212	44	that	that	PRON
fuelectenerg-13681	212	45	indicates	indicate	VERB
fuelectenerg-13681	212	46	ill	ill	ADJ
fuelectenerg-13681	212	47	-	-	PUNCT
fuelectenerg-13681	212	48	conditioning	conditioning	NOUN
fuelectenerg-13681	212	49	,	,	PUNCT
fuelectenerg-13681	212	50	where	where	SCONJ
fuelectenerg-13681	212	51	traditional	traditional	ADJ
fuelectenerg-13681	212	52	methods	method	NOUN
fuelectenerg-13681	212	53	may	may	AUX
fuelectenerg-13681	212	54	struggle	struggle	VERB
fuelectenerg-13681	212	55	;	;	PUNCT
fuelectenerg-13681	212	56	interference	interference	NOUN
fuelectenerg-13681	212	57	level	level	NOUN
fuelectenerg-13681	212	58	that	that	PRON
fuelectenerg-13681	212	59	measures	measure	VERB
fuelectenerg-13681	212	60	the	the	DET
fuelectenerg-13681	212	61	ratio	ratio	NOUN
fuelectenerg-13681	212	62	of	of	ADP
fuelectenerg-13681	212	63	interference	interference	NOUN
fuelectenerg-13681	212	64	to	to	PART
fuelectenerg-13681	212	65	signal	signal	VERB
fuelectenerg-13681	212	66	power	power	NOUN
fuelectenerg-13681	212	67	;	;	PUNCT
fuelectenerg-13681	212	68	and	and	CCONJ
fuelectenerg-13681	212	69	computational	computational	ADJ
fuelectenerg-13681	212	70	budget	budget	NOUN
fuelectenerg-13681	212	71	that	that	PRON
fuelectenerg-13681	212	72	ensures	ensure	VERB
fuelectenerg-13681	212	73	the	the	DET
fuelectenerg-13681	212	74	selected	select	VERB
fuelectenerg-13681	212	75	algorithm	algorithm	NOUN
fuelectenerg-13681	212	76	can	can	AUX
fuelectenerg-13681	212	77	execute	execute	VERB
fuelectenerg-13681	212	78	within	within	ADP
fuelectenerg-13681	212	79	time	time	NOUN
fuelectenerg-13681	212	80	constraints	constraint	NOUN
fuelectenerg-13681	212	81	.	.	PUNCT
fuelectenerg-13681	213	1	we	we	PRON
fuelectenerg-13681	213	2	prove	prove	VERB
fuelectenerg-13681	213	3	that	that	SCONJ
fuelectenerg-13681	213	4	under	under	ADP
fuelectenerg-13681	213	5	mild	mild	ADJ
fuelectenerg-13681	213	6	conditions	condition	NOUN
fuelectenerg-13681	213	7	on	on	ADP
fuelectenerg-13681	213	8	the	the	DET
fuelectenerg-13681	213	9	reward	reward	NOUN
fuelectenerg-13681	213	10	distributions	distribution	NOUN
fuelectenerg-13681	213	11	,	,	PUNCT
fuelectenerg-13681	213	12	the	the	DET
fuelectenerg-13681	213	13	regret	regret	NOUN
fuelectenerg-13681	213	14	of	of	ADP
fuelectenerg-13681	213	15	this	this	DET
fuelectenerg-13681	213	16	selection	selection	NOUN
fuelectenerg-13681	213	17	mechanism	mechanism	NOUN
fuelectenerg-13681	213	18	is	be	AUX
fuelectenerg-13681	213	19	bounded	bound	VERB
fuelectenerg-13681	213	20	by	by	ADP
fuelectenerg-13681	213	21	(	(	PUNCT
fuelectenerg-13681	213	22	log	log	PROPN
fuelectenerg-13681	213	23	)	)	PUNCT
fuelectenerg-13681	214	1	o	o	NOUN
fuelectenerg-13681	214	2	t	t	PROPN
fuelectenerg-13681	214	3	t	t	X
fuelectenerg-13681	214	4	where	where	SCONJ
fuelectenerg-13681	214	5	t	t	PROPN
fuelectenerg-13681	214	6	is	be	AUX
fuelectenerg-13681	214	7	the	the	DET
fuelectenerg-13681	214	8	number	number	NOUN
fuelectenerg-13681	214	9	of	of	ADP
fuelectenerg-13681	214	10	time	time	NOUN
fuelectenerg-13681	214	11	steps	step	NOUN
fuelectenerg-13681	214	12	.	.	PUNCT
fuelectenerg-13681	215	1	this	this	PRON
fuelectenerg-13681	215	2	means	mean	VERB
fuelectenerg-13681	215	3	that	that	SCONJ
fuelectenerg-13681	215	4	the	the	DET
fuelectenerg-13681	215	5	performance	performance	NOUN
fuelectenerg-13681	215	6	of	of	ADP
fuelectenerg-13681	215	7	our	our	PRON
fuelectenerg-13681	215	8	hybrid	hybrid	ADJ
fuelectenerg-13681	215	9	approach	approach	NOUN
fuelectenerg-13681	215	10	asymptotically	asymptotically	ADV
fuelectenerg-13681	215	11	approaches	approach	VERB
fuelectenerg-13681	215	12	that	that	PRON
fuelectenerg-13681	215	13	of	of	ADP
fuelectenerg-13681	215	14	the	the	DET
fuelectenerg-13681	215	15	optimal	optimal	ADJ
fuelectenerg-13681	215	16	algorithm	algorithm	NOUN
fuelectenerg-13681	215	17	selection	selection	NOUN
fuelectenerg-13681	215	18	strategy	strategy	NOUN
fuelectenerg-13681	215	19	.	.	PUNCT
fuelectenerg-13681	216	1	3.5	3.5	NUM
fuelectenerg-13681	216	2	.	.	PUNCT
fuelectenerg-13681	217	1	datasets	dataset	NOUN
fuelectenerg-13681	217	2	we	we	PRON
fuelectenerg-13681	217	3	utilize	utilize	VERB
fuelectenerg-13681	217	4	three	three	NUM
fuelectenerg-13681	217	5	comprehensive	comprehensive	ADJ
fuelectenerg-13681	217	6	open	open	ADJ
fuelectenerg-13681	217	7	-	-	PUNCT
fuelectenerg-13681	217	8	source	source	NOUN
fuelectenerg-13681	217	9	datasets	dataset	NOUN
fuelectenerg-13681	217	10	for	for	ADP
fuelectenerg-13681	217	11	our	our	PRON
fuelectenerg-13681	217	12	evaluation	evaluation	NOUN
fuelectenerg-13681	217	13	,	,	PUNCT
fuelectenerg-13681	217	14	each	each	PRON
fuelectenerg-13681	217	15	designed	design	VERB
fuelectenerg-13681	217	16	to	to	PART
fuelectenerg-13681	217	17	capture	capture	VERB
fuelectenerg-13681	217	18	different	different	ADJ
fuelectenerg-13681	217	19	aspects	aspect	NOUN
fuelectenerg-13681	217	20	of	of	ADP
fuelectenerg-13681	217	21	mimo	mimo	PROPN
fuelectenerg-13681	217	22	channel	channel	PROPN
fuelectenerg-13681	217	23	conditions	condition	NOUN
fuelectenerg-13681	217	24	:	:	PUNCT
fuelectenerg-13681	217	25	deepmimo	deepmimo	VERB
fuelectenerg-13681	217	26	:	:	PUNCT
fuelectenerg-13681	217	27	this	this	DET
fuelectenerg-13681	217	28	publicly	publicly	ADV
fuelectenerg-13681	217	29	available	available	ADJ
fuelectenerg-13681	217	30	dataset	dataset	NOUN
fuelectenerg-13681	218	1	[	[	X
fuelectenerg-13681	218	2	35	35	NUM
fuelectenerg-13681	218	3	]	]	PUNCT
fuelectenerg-13681	218	4	provides	provide	VERB
fuelectenerg-13681	218	5	realistic	realistic	ADJ
fuelectenerg-13681	218	6	mimo	mimo	PROPN
fuelectenerg-13681	218	7	channel	channel	PROPN
fuelectenerg-13681	218	8	realizations	realization	NOUN
fuelectenerg-13681	218	9	based	base	VERB
fuelectenerg-13681	218	10	on	on	ADP
fuelectenerg-13681	218	11	ray	ray	NOUN
fuelectenerg-13681	218	12	-	-	PUNCT
fuelectenerg-13681	218	13	tracing	tracing	NOUN
fuelectenerg-13681	218	14	.	.	PUNCT
fuelectenerg-13681	219	1	we	we	PRON
fuelectenerg-13681	219	2	utilize	utilize	VERB
fuelectenerg-13681	219	3	scenarios	scenario	NOUN
fuelectenerg-13681	219	4	’	'	PUNCT
fuelectenerg-13681	219	5	o1	o1	NOUN
fuelectenerg-13681	219	6	’	'	PUNCT
fuelectenerg-13681	219	7	(	(	PUNCT
fuelectenerg-13681	219	8	outdoor	outdoor	NOUN
fuelectenerg-13681	219	9	)	)	PUNCT
fuelectenerg-13681	219	10	,	,	PUNCT
fuelectenerg-13681	219	11	’	'	PUNCT
fuelectenerg-13681	219	12	i1	i1	PROPN
fuelectenerg-13681	219	13	’	'	PUNCT
fuelectenerg-13681	219	14	(	(	PUNCT
fuelectenerg-13681	219	15	indoor	indoor	ADJ
fuelectenerg-13681	219	16	office	office	NOUN
fuelectenerg-13681	219	17	)	)	PUNCT
fuelectenerg-13681	219	18	,	,	PUNCT
fuelectenerg-13681	219	19	and	and	CCONJ
fuelectenerg-13681	219	20	’	'	PUNCT
fuelectenerg-13681	219	21	i3	i3	NOUN
fuelectenerg-13681	219	22	’	'	PUNCT
fuelectenerg-13681	219	23	(	(	PUNCT
fuelectenerg-13681	219	24	indoor	indoor	ADJ
fuelectenerg-13681	219	25	shopping	shopping	NOUN
fuelectenerg-13681	219	26	mall	mall	NOUN
fuelectenerg-13681	219	27	)	)	PUNCT
fuelectenerg-13681	219	28	,	,	PUNCT
fuelectenerg-13681	219	29	comprising	comprise	VERB
fuelectenerg-13681	219	30	approximately	approximately	ADV
fuelectenerg-13681	219	31	50,000	50,000	NUM
fuelectenerg-13681	219	32	channel	channel	NOUN
fuelectenerg-13681	219	33	realizations	realization	NOUN
fuelectenerg-13681	219	34	.	.	PUNCT
fuelectenerg-13681	220	1	the	the	DET
fuelectenerg-13681	220	2	deep	deep	ADJ
fuelectenerg-13681	220	3	learning	learning	NOUN
fuelectenerg-13681	220	4	and	and	CCONJ
fuelectenerg-13681	220	5	traditional	traditional	ADJ
fuelectenerg-13681	220	6	algorithms	algorithm	NOUN
fuelectenerg-13681	220	7	for	for	ADP
fuelectenerg-13681	220	8	mimo	mimo	PROPN
fuelectenerg-13681	220	9	beamforming	beamforme	VERB
fuelectenerg-13681	220	10	207	207	NUM
fuelectenerg-13681	220	11	dataset	dataset	NOUN
fuelectenerg-13681	220	12	includes	include	VERB
fuelectenerg-13681	220	13	mimo	mimo	PROPN
fuelectenerg-13681	220	14	configurations	configuration	NOUN
fuelectenerg-13681	220	15	ranging	range	VERB
fuelectenerg-13681	220	16	from	from	ADP
fuelectenerg-13681	220	17	4×4	4×4	NUM
fuelectenerg-13681	220	18	to	to	ADP
fuelectenerg-13681	220	19	64×64	64×64	NUM
fuelectenerg-13681	220	20	,	,	PUNCT
fuelectenerg-13681	220	21	with	with	ADP
fuelectenerg-13681	220	22	mobility	mobility	NOUN
fuelectenerg-13681	220	23	scenarios	scenario	NOUN
fuelectenerg-13681	220	24	from	from	ADP
fuelectenerg-13681	220	25	static	static	ADJ
fuelectenerg-13681	220	26	to	to	ADP
fuelectenerg-13681	220	27	high	high	ADJ
fuelectenerg-13681	220	28	-	-	PUNCT
fuelectenerg-13681	220	29	speed	speed	NOUN
fuelectenerg-13681	220	30	(	(	PUNCT
fuelectenerg-13681	220	31	300	300	NUM
fuelectenerg-13681	220	32	km	km	NOUN
fuelectenerg-13681	220	33	/	/	SYM
fuelectenerg-13681	220	34	h	h	NOUN
fuelectenerg-13681	220	35	)	)	PUNCT
fuelectenerg-13681	220	36	.	.	PUNCT
fuelectenerg-13681	221	1	3gpp	3gpp	NOUN
fuelectenerg-13681	221	2	tr	tr	PUNCT
fuelectenerg-13681	221	3	38.901	38.901	NUM
fuelectenerg-13681	221	4	channel	channel	NOUN
fuelectenerg-13681	221	5	model	model	NOUN
fuelectenerg-13681	221	6	:	:	PUNCT
fuelectenerg-13681	221	7	we	we	PRON
fuelectenerg-13681	221	8	generate	generate	VERB
fuelectenerg-13681	221	9	30,000	30,000	NUM
fuelectenerg-13681	221	10	channel	channel	NOUN
fuelectenerg-13681	221	11	realizations	realization	NOUN
fuelectenerg-13681	221	12	using	use	VERB
fuelectenerg-13681	221	13	the	the	DET
fuelectenerg-13681	221	14	standardized	standardized	ADJ
fuelectenerg-13681	221	15	3gpp	3gpp	PROPN
fuelectenerg-13681	221	16	tr	tr	VERB
fuelectenerg-13681	221	17	38.901	38.901	NUM
fuelectenerg-13681	221	18	channel	channel	NOUN
fuelectenerg-13681	221	19	model	model	NOUN
fuelectenerg-13681	221	20	[	[	X
fuelectenerg-13681	221	21	36	36	NUM
fuelectenerg-13681	221	22	]	]	PUNCT
fuelectenerg-13681	221	23	.	.	PUNCT
fuelectenerg-13681	222	1	this	this	PRON
fuelectenerg-13681	222	2	includes	include	VERB
fuelectenerg-13681	222	3	urban	urban	ADJ
fuelectenerg-13681	222	4	macro	macro	NOUN
fuelectenerg-13681	222	5	(	(	PUNCT
fuelectenerg-13681	222	6	uma	uma	PROPN
fuelectenerg-13681	222	7	)	)	PUNCT
fuelectenerg-13681	222	8	,	,	PUNCT
fuelectenerg-13681	222	9	urban	urban	PROPN
fuelectenerg-13681	222	10	micro	micro	PROPN
fuelectenerg-13681	222	11	(	(	PUNCT
fuelectenerg-13681	222	12	umi	umi	PROPN
fuelectenerg-13681	222	13	)	)	PUNCT
fuelectenerg-13681	222	14	,	,	PUNCT
fuelectenerg-13681	222	15	and	and	CCONJ
fuelectenerg-13681	222	16	rural	rural	ADJ
fuelectenerg-13681	222	17	macro	macro	NOUN
fuelectenerg-13681	222	18	(	(	PUNCT
fuelectenerg-13681	222	19	rma	rma	NOUN
fuelectenerg-13681	222	20	)	)	PUNCT
fuelectenerg-13681	222	21	scenarios	scenario	NOUN
fuelectenerg-13681	222	22	with	with	ADP
fuelectenerg-13681	222	23	various	various	ADJ
fuelectenerg-13681	222	24	user	user	NOUN
fuelectenerg-13681	222	25	speeds	speed	NOUN
fuelectenerg-13681	222	26	from	from	ADP
fuelectenerg-13681	222	27	3	3	NUM
fuelectenerg-13681	222	28	km	km	NOUN
fuelectenerg-13681	222	29	/	/	SYM
fuelectenerg-13681	222	30	h	h	NOUN
fuelectenerg-13681	222	31	to	to	ADP
fuelectenerg-13681	222	32	120	120	NUM
fuelectenerg-13681	222	33	km	km	PROPN
fuelectenerg-13681	222	34	/	/	SYM
fuelectenerg-13681	222	35	h.	h.	PROPN
fuelectenerg-13681	223	1	the	the	DET
fuelectenerg-13681	223	2	channel	channel	NOUN
fuelectenerg-13681	223	3	matrices	matrix	NOUN
fuelectenerg-13681	223	4	incorporate	incorporate	VERB
fuelectenerg-13681	223	5	realistic	realistic	ADJ
fuelectenerg-13681	223	6	path	path	NOUN
fuelectenerg-13681	223	7	loss	loss	NOUN
fuelectenerg-13681	223	8	,	,	PUNCT
fuelectenerg-13681	223	9	shadowing	shadowing	NOUN
fuelectenerg-13681	223	10	,	,	PUNCT
fuelectenerg-13681	223	11	and	and	CCONJ
fuelectenerg-13681	223	12	small	small	ADJ
fuelectenerg-13681	223	13	-	-	PUNCT
fuelectenerg-13681	223	14	scale	scale	NOUN
fuelectenerg-13681	223	15	fading	fading	NOUN
fuelectenerg-13681	223	16	effects	effect	NOUN
fuelectenerg-13681	223	17	following	follow	VERB
fuelectenerg-13681	223	18	the	the	DET
fuelectenerg-13681	223	19	3gpp	3gpp	PROPN
fuelectenerg-13681	223	20	specifications	specification	NOUN
fuelectenerg-13681	223	21	.	.	PUNCT
fuelectenerg-13681	224	1	ieee	ieee	PROPN
fuelectenerg-13681	224	2	mimo	mimo	PROPN
fuelectenerg-13681	224	3	data	data	PROPN
fuelectenerg-13681	224	4	challenge	challenge	NOUN
fuelectenerg-13681	224	5	:	:	PUNCT
fuelectenerg-13681	224	6	this	this	PRON
fuelectenerg-13681	224	7	dataset	dataset	VERB
fuelectenerg-13681	224	8	[	[	X
fuelectenerg-13681	224	9	37	37	NUM
fuelectenerg-13681	224	10	]	]	PUNCT
fuelectenerg-13681	224	11	contains	contain	VERB
fuelectenerg-13681	224	12	real	real	ADJ
fuelectenerg-13681	224	13	-	-	PUNCT
fuelectenerg-13681	224	14	world	world	NOUN
fuelectenerg-13681	224	15	channel	channel	NOUN
fuelectenerg-13681	224	16	measurements	measurement	NOUN
fuelectenerg-13681	224	17	from	from	ADP
fuelectenerg-13681	224	18	the	the	DET
fuelectenerg-13681	224	19	ieee	ieee	PROPN
fuelectenerg-13681	224	20	mimo	mimo	PROPN
fuelectenerg-13681	224	21	data	data	PROPN
fuelectenerg-13681	224	22	challenge	challenge	NOUN
fuelectenerg-13681	224	23	,	,	PUNCT
fuelectenerg-13681	224	24	including	include	VERB
fuelectenerg-13681	224	25	both	both	CCONJ
fuelectenerg-13681	224	26	indoor	indoor	ADJ
fuelectenerg-13681	224	27	and	and	CCONJ
fuelectenerg-13681	224	28	outdoor	outdoor	ADJ
fuelectenerg-13681	224	29	scenarios	scenario	NOUN
fuelectenerg-13681	224	30	.	.	PUNCT
fuelectenerg-13681	225	1	the	the	DET
fuelectenerg-13681	225	2	dataset	dataset	NOUN
fuelectenerg-13681	225	3	provides	provide	VERB
fuelectenerg-13681	225	4	approximately	approximately	ADV
fuelectenerg-13681	225	5	20,000	20,000	NUM
fuelectenerg-13681	225	6	channel	channel	NOUN
fuelectenerg-13681	225	7	matrices	matrix	NOUN
fuelectenerg-13681	225	8	measured	measure	VERB
fuelectenerg-13681	225	9	in	in	ADP
fuelectenerg-13681	225	10	diverse	diverse	ADJ
fuelectenerg-13681	225	11	environments	environment	NOUN
fuelectenerg-13681	225	12	with	with	ADP
fuelectenerg-13681	225	13	various	various	ADJ
fuelectenerg-13681	225	14	antenna	antenna	NOUN
fuelectenerg-13681	225	15	configurations	configuration	NOUN
fuelectenerg-13681	225	16	.	.	PUNCT
fuelectenerg-13681	226	1	we	we	PRON
fuelectenerg-13681	226	2	particularly	particularly	ADV
fuelectenerg-13681	226	3	focus	focus	VERB
fuelectenerg-13681	226	4	on	on	ADP
fuelectenerg-13681	226	5	the	the	DET
fuelectenerg-13681	226	6	high	high	ADJ
fuelectenerg-13681	226	7	-	-	PUNCT
fuelectenerg-13681	226	8	mobility	mobility	NOUN
fuelectenerg-13681	226	9	scenarios	scenario	NOUN
fuelectenerg-13681	226	10	and	and	CCONJ
fuelectenerg-13681	226	11	those	those	PRON
fuelectenerg-13681	226	12	with	with	ADP
fuelectenerg-13681	226	13	significant	significant	ADJ
fuelectenerg-13681	226	14	multipath	multipath	NOUN
fuelectenerg-13681	226	15	effects	effect	NOUN
fuelectenerg-13681	226	16	.	.	PUNCT
fuelectenerg-13681	227	1	we	we	PRON
fuelectenerg-13681	227	2	validated	validate	VERB
fuelectenerg-13681	227	3	the	the	DET
fuelectenerg-13681	227	4	statistical	statistical	ADJ
fuelectenerg-13681	227	5	properties	property	NOUN
fuelectenerg-13681	227	6	of	of	ADP
fuelectenerg-13681	227	7	our	our	PRON
fuelectenerg-13681	227	8	datasets	dataset	NOUN
fuelectenerg-13681	227	9	against	against	ADP
fuelectenerg-13681	227	10	theoretical	theoretical	ADJ
fuelectenerg-13681	227	11	channel	channel	NOUN
fuelectenerg-13681	227	12	models	model	NOUN
fuelectenerg-13681	227	13	and	and	CCONJ
fuelectenerg-13681	227	14	real	real	ADJ
fuelectenerg-13681	227	15	-	-	PUNCT
fuelectenerg-13681	227	16	world	world	NOUN
fuelectenerg-13681	227	17	measurements	measurement	NOUN
fuelectenerg-13681	227	18	.	.	PUNCT
fuelectenerg-13681	228	1	the	the	DET
fuelectenerg-13681	228	2	empirical	empirical	ADJ
fuelectenerg-13681	228	3	eigenvalue	eigenvalue	NOUN
fuelectenerg-13681	228	4	distributions	distribution	NOUN
fuelectenerg-13681	228	5	of	of	ADP
fuelectenerg-13681	228	6	the	the	DET
fuelectenerg-13681	228	7	channel	channel	NOUN
fuelectenerg-13681	228	8	correlation	correlation	NOUN
fuelectenerg-13681	228	9	matrices	matrix	NOUN
fuelectenerg-13681	228	10	confirm	confirm	VERB
fuelectenerg-13681	228	11	that	that	SCONJ
fuelectenerg-13681	228	12	these	these	DET
fuelectenerg-13681	228	13	datasets	dataset	NOUN
fuelectenerg-13681	228	14	accurately	accurately	ADV
fuelectenerg-13681	228	15	represent	represent	VERB
fuelectenerg-13681	228	16	realistic	realistic	ADJ
fuelectenerg-13681	228	17	mimo	mimo	PROPN
fuelectenerg-13681	228	18	channels	channel	NOUN
fuelectenerg-13681	228	19	.	.	PUNCT
fuelectenerg-13681	229	1	3.6	3.6	NUM
fuelectenerg-13681	229	2	.	.	PUNCT
fuelectenerg-13681	229	3	evaluation	evaluation	NOUN
fuelectenerg-13681	229	4	metrics	metric	NOUN
fuelectenerg-13681	229	5	we	we	PRON
fuelectenerg-13681	229	6	evaluate	evaluate	VERB
fuelectenerg-13681	229	7	the	the	DET
fuelectenerg-13681	229	8	performance	performance	NOUN
fuelectenerg-13681	229	9	of	of	ADP
fuelectenerg-13681	229	10	both	both	CCONJ
fuelectenerg-13681	229	11	traditional	traditional	ADJ
fuelectenerg-13681	229	12	and	and	CCONJ
fuelectenerg-13681	229	13	deep	deep	ADJ
fuelectenerg-13681	229	14	learning	learning	NOUN
fuelectenerg-13681	229	15	approaches	approach	NOUN
fuelectenerg-13681	229	16	using	use	VERB
fuelectenerg-13681	229	17	the	the	DET
fuelectenerg-13681	229	18	following	follow	VERB
fuelectenerg-13681	229	19	comprehensive	comprehensive	ADJ
fuelectenerg-13681	229	20	set	set	NOUN
fuelectenerg-13681	229	21	of	of	ADP
fuelectenerg-13681	229	22	metrics	metric	NOUN
fuelectenerg-13681	229	23	:	:	PUNCT
fuelectenerg-13681	229	24	▪	▪	ADJ
fuelectenerg-13681	229	25	signal	signal	NOUN
fuelectenerg-13681	229	26	-	-	PUNCT
fuelectenerg-13681	229	27	to	to	ADP
fuelectenerg-13681	229	28	-	-	PUNCT
fuelectenerg-13681	229	29	interference	interference	NOUN
fuelectenerg-13681	229	30	-	-	PUNCT
fuelectenerg-13681	229	31	plus	plus	CCONJ
fuelectenerg-13681	229	32	-	-	PUNCT
fuelectenerg-13681	229	33	noise	noise	NOUN
fuelectenerg-13681	229	34	ratio	ratio	NOUN
fuelectenerg-13681	229	35	(	(	PUNCT
fuelectenerg-13681	229	36	sinr	sinr	INTJ
fuelectenerg-13681	229	37	):	):	PUNCT
fuelectenerg-13681	229	38	2	2	NUM
fuelectenerg-13681	229	39	signal	signal	NOUN
fuelectenerg-13681	229	40	2	2	NUM
fuelectenerg-13681	229	41	2	2	NUM
fuelectenerg-13681	229	42	interference	interference	NOUN
fuelectenerg-13681	229	43	noise	noise	NOUN
fuelectenerg-13681	230	1	|	|	ADV
fuelectenerg-13681	230	2	|	|	ADV
fuelectenerg-13681	230	3	sinr	sinr	VERB
fuelectenerg-13681	230	4	‍|	‍|	PROPN
fuelectenerg-13681	231	1	|	|	NOUN
fuelectenerg-13681	232	1	d	d	NOUN
fuelectenerg-13681	232	2	i	i	PRON
fuelectenerg-13681	233	1	i	i	VERB
fuelectenerg-13681	234	1	d	d	X
fuelectenerg-13681	234	2	p	p	X
fuelectenerg-13681	234	3	p	p	X
fuelectenerg-13681	234	4	p	p	X
fuelectenerg-13681	234	5			PROPN
fuelectenerg-13681	234	6			NOUN
fuelectenerg-13681	234	7	=	=	PUNCT
fuelectenerg-13681	235	1	=	=	PUNCT
fuelectenerg-13681	236	1	+	+	PUNCT
fuelectenerg-13681	236	2	+	+	X
fuelectenerg-13681	236	3			PART
fuelectenerg-13681	236	4	hwx	hwx	VERB
fuelectenerg-13681	236	5	hwx	hwx	NOUN
fuelectenerg-13681	236	6	(	(	PUNCT
fuelectenerg-13681	236	7	33	33	NUM
fuelectenerg-13681	236	8	)	)	PUNCT
fuelectenerg-13681	236	9	where	where	SCONJ
fuelectenerg-13681	236	10	dx	dx	PROPN
fuelectenerg-13681	236	11	is	be	AUX
fuelectenerg-13681	236	12	the	the	DET
fuelectenerg-13681	236	13	desired	desire	VERB
fuelectenerg-13681	236	14	signal	signal	NOUN
fuelectenerg-13681	236	15	and	and	CCONJ
fuelectenerg-13681	236	16	ix	ix	ADV
fuelectenerg-13681	236	17	represents	represent	VERB
fuelectenerg-13681	236	18	interfering	interfere	VERB
fuelectenerg-13681	236	19	signals	signal	NOUN
fuelectenerg-13681	236	20	.	.	PUNCT
fuelectenerg-13681	237	1	▪	▪	ADJ
fuelectenerg-13681	237	2	bit	bit	NOUN
fuelectenerg-13681	237	3	error	error	NOUN
fuelectenerg-13681	237	4	rate	rate	NOUN
fuelectenerg-13681	237	5	(	(	PUNCT
fuelectenerg-13681	237	6	ber	ber	PROPN
fuelectenerg-13681	237	7	):	):	PUNCT
fuelectenerg-13681	237	8	numberofbiterrors	numberofbiterror	NOUN
fuelectenerg-13681	237	9	ber	ber	NOUN
fuelectenerg-13681	237	10	totalnumberoftransmittedbits	totalnumberoftransmittedbit	NOUN
fuelectenerg-13681	237	11	=	=	SYM
fuelectenerg-13681	237	12	(	(	PUNCT
fuelectenerg-13681	237	13	34	34	NUM
fuelectenerg-13681	237	14	)	)	PUNCT
fuelectenerg-13681	237	15	▪	▪	NOUN
fuelectenerg-13681	237	16	spectral	spectral	ADJ
fuelectenerg-13681	237	17	efficiency	efficiency	NOUN
fuelectenerg-13681	237	18	(	(	PUNCT
fuelectenerg-13681	237	19	bits	bit	NOUN
fuelectenerg-13681	237	20	/	/	SYM
fuelectenerg-13681	237	21	s	s	PROPN
fuelectenerg-13681	237	22	/	/	SYM
fuelectenerg-13681	237	23	hz	hz	ADJ
fuelectenerg-13681	237	24	):	):	PUNCT
fuelectenerg-13681	237	25	2	2	NUM
fuelectenerg-13681	237	26	(	(	PUNCT
fuelectenerg-13681	237	27	)	)	PUNCT
fuelectenerg-13681	237	28	se	se	X
fuelectenerg-13681	237	29	log	log	VERB
fuelectenerg-13681	237	30	1	1	NUM
fuelectenerg-13681	237	31	sinr=	sinr=	NUM
fuelectenerg-13681	237	32	+	+	CCONJ
fuelectenerg-13681	237	33	(	(	PUNCT
fuelectenerg-13681	237	34	35	35	NUM
fuelectenerg-13681	237	35	)	)	PUNCT
fuelectenerg-13681	237	36	▪	▪	NOUN
fuelectenerg-13681	237	37	computational	computational	ADJ
fuelectenerg-13681	237	38	complexity	complexity	NOUN
fuelectenerg-13681	237	39	:	:	PUNCT
fuelectenerg-13681	237	40	measured	measure	VERB
fuelectenerg-13681	237	41	in	in	ADP
fuelectenerg-13681	237	42	floating	float	VERB
fuelectenerg-13681	237	43	-	-	PUNCT
fuelectenerg-13681	237	44	point	point	NOUN
fuelectenerg-13681	237	45	operations	operation	NOUN
fuelectenerg-13681	237	46	(	(	PUNCT
fuelectenerg-13681	237	47	flops	flop	NOUN
fuelectenerg-13681	237	48	)	)	PUNCT
fuelectenerg-13681	237	49	and	and	CCONJ
fuelectenerg-13681	237	50	execution	execution	NOUN
fuelectenerg-13681	237	51	time	time	NOUN
fuelectenerg-13681	237	52	(	(	PUNCT
fuelectenerg-13681	237	53	ms	ms	NOUN
fuelectenerg-13681	237	54	)	)	PUNCT
fuelectenerg-13681	237	55	required	require	VERB
fuelectenerg-13681	237	56	to	to	PART
fuelectenerg-13681	237	57	compute	compute	VERB
fuelectenerg-13681	237	58	the	the	DET
fuelectenerg-13681	237	59	beamforming	beamforme	VERB
fuelectenerg-13681	237	60	weights	weight	NOUN
fuelectenerg-13681	237	61	.	.	PUNCT
fuelectenerg-13681	238	1	▪	▪	ADJ
fuelectenerg-13681	238	2	convergence	convergence	NOUN
fuelectenerg-13681	238	3	speed	speed	NOUN
fuelectenerg-13681	238	4	:	:	PUNCT
fuelectenerg-13681	238	5	measured	measure	VERB
fuelectenerg-13681	238	6	as	as	ADP
fuelectenerg-13681	238	7	the	the	DET
fuelectenerg-13681	238	8	number	number	NOUN
fuelectenerg-13681	238	9	of	of	ADP
fuelectenerg-13681	238	10	iterations	iteration	NOUN
fuelectenerg-13681	238	11	or	or	CCONJ
fuelectenerg-13681	238	12	samples	sample	NOUN
fuelectenerg-13681	238	13	required	require	VERB
fuelectenerg-13681	238	14	to	to	PART
fuelectenerg-13681	238	15	achieve	achieve	VERB
fuelectenerg-13681	238	16	within	within	ADP
fuelectenerg-13681	238	17	5	5	NUM
fuelectenerg-13681	238	18	%	%	NOUN
fuelectenerg-13681	238	19	of	of	ADP
fuelectenerg-13681	238	20	the	the	DET
fuelectenerg-13681	238	21	optimal	optimal	ADJ
fuelectenerg-13681	238	22	sinr	sinr	NOUN
fuelectenerg-13681	238	23	.	.	PUNCT
fuelectenerg-13681	239	1	optimalconvergencetime	optimalconvergencetime	PROPN
fuelectenerg-13681	239	2	min	min	PROPN
fuelectenerg-13681	239	3	{	{	PUNCT
fuelectenerg-13681	239	4	:	:	PUNCT
fuelectenerg-13681	239	5	sinr	sinr	NOUN
fuelectenerg-13681	239	6	(	(	PUNCT
fuelectenerg-13681	239	7	)	)	PUNCT
fuelectenerg-13681	239	8	0.95	0.95	NUM
fuelectenerg-13681	239	9	sinr	sinr	NOUN
fuelectenerg-13681	239	10	}	}	PUNCT
fuelectenerg-13681	239	11	t	t	NOUN
fuelectenerg-13681	239	12	t=	t=	NOUN
fuelectenerg-13681	239	13			NUM
fuelectenerg-13681	239	14			PROPN
fuelectenerg-13681	239	15	(	(	PUNCT
fuelectenerg-13681	239	16	36	36	NUM
fuelectenerg-13681	239	17	)	)	PUNCT
fuelectenerg-13681	239	18	▪	▪	NOUN
fuelectenerg-13681	239	19	adaptability	adaptability	NOUN
fuelectenerg-13681	239	20	:	:	PUNCT
fuelectenerg-13681	239	21	quantifies	quantify	VERB
fuelectenerg-13681	239	22	an	an	DET
fuelectenerg-13681	239	23	algorithm	algorithm	NOUN
fuelectenerg-13681	239	24	’s	’s	PART
fuelectenerg-13681	239	25	ability	ability	NOUN
fuelectenerg-13681	239	26	to	to	PART
fuelectenerg-13681	239	27	maintain	maintain	VERB
fuelectenerg-13681	239	28	performance	performance	NOUN
fuelectenerg-13681	239	29	in	in	ADP
fuelectenerg-13681	239	30	changing	change	VERB
fuelectenerg-13681	239	31	channel	channel	NOUN
fuelectenerg-13681	239	32	conditions	condition	NOUN
fuelectenerg-13681	239	33	,	,	PUNCT
fuelectenerg-13681	239	34	defined	define	VERB
fuelectenerg-13681	239	35	as	as	ADP
fuelectenerg-13681	239	36	:	:	PUNCT
fuelectenerg-13681	239	37	optimal,1	optimal,1	ADJ
fuelectenerg-13681	239	38	sinr1	sinr1	PROPN
fuelectenerg-13681	239	39	adaptability	adaptability	NOUN
fuelectenerg-13681	239	40	‍	‍	PROPN
fuelectenerg-13681	239	41	sinr	sinr	INTJ
fuelectenerg-13681	239	42	t	t	PROPN
fuelectenerg-13681	239	43	t	t	PROPN
fuelectenerg-13681	239	44	tt	tt	X
fuelectenerg-13681	239	45	t	t	PROPN
fuelectenerg-13681	239	46	=	=	PUNCT
fuelectenerg-13681	240	1	=	=	SYM
fuelectenerg-13681	240	2			X
fuelectenerg-13681	240	3	(	(	PUNCT
fuelectenerg-13681	240	4	37	37	NUM
fuelectenerg-13681	240	5	)	)	SYM
fuelectenerg-13681	240	6	208	208	NUM
fuelectenerg-13681	240	7	m.	m.	NOUN
fuelectenerg-13681	240	8	mohamed	mohamed	PROPN
fuelectenerg-13681	240	9	,	,	PUNCT
fuelectenerg-13681	240	10	f.	f.	PROPN
fuelectenerg-13681	240	11	aljuaid	aljuaid	PROPN
fuelectenerg-13681	240	12	,	,	PUNCT
fuelectenerg-13681	240	13	mhd	mhd	PROPN
fuelectenerg-13681	240	14	w.	w.	PROPN
fuelectenerg-13681	240	15	koubeisi	koubeisi	VERB
fuelectenerg-13681	240	16	where	where	SCONJ
fuelectenerg-13681	240	17	sinrt	sinrt	NOUN
fuelectenerg-13681	240	18	is	be	AUX
fuelectenerg-13681	240	19	the	the	DET
fuelectenerg-13681	240	20	achieved	achieve	VERB
fuelectenerg-13681	240	21	sinr	sinr	NOUN
fuelectenerg-13681	240	22	at	at	ADP
fuelectenerg-13681	240	23	time	time	NOUN
fuelectenerg-13681	240	24	t	t	PROPN
fuelectenerg-13681	240	25	sinroptimal	sinroptimal	NOUN
fuelectenerg-13681	240	26	,	,	PUNCT
fuelectenerg-13681	240	27	t	t	PROPN
fuelectenerg-13681	240	28	is	be	AUX
fuelectenerg-13681	240	29	the	the	DET
fuelectenerg-13681	240	30	theoretical	theoretical	ADJ
fuelectenerg-13681	240	31	optimal	optimal	ADJ
fuelectenerg-13681	240	32	sinr	sinr	NOUN
fuelectenerg-13681	240	33	at	at	ADP
fuelectenerg-13681	240	34	time	time	NOUN
fuelectenerg-13681	240	35	t	t	PROPN
fuelectenerg-13681	240	36	and	and	CCONJ
fuelectenerg-13681	240	37	t	t	PROPN
fuelectenerg-13681	240	38	is	be	AUX
fuelectenerg-13681	240	39	the	the	DET
fuelectenerg-13681	240	40	number	number	NOUN
fuelectenerg-13681	240	41	of	of	ADP
fuelectenerg-13681	240	42	time	time	NOUN
fuelectenerg-13681	240	43	steps	step	NOUN
fuelectenerg-13681	240	44	in	in	ADP
fuelectenerg-13681	240	45	the	the	DET
fuelectenerg-13681	240	46	evaluation	evaluation	NOUN
fuelectenerg-13681	240	47	period	period	NOUN
fuelectenerg-13681	240	48	.	.	PUNCT
fuelectenerg-13681	241	1	higher	high	ADJ
fuelectenerg-13681	241	2	values	value	NOUN
fuelectenerg-13681	241	3	indicate	indicate	VERB
fuelectenerg-13681	241	4	better	well	ADJ
fuelectenerg-13681	241	5	adaptability	adaptability	NOUN
fuelectenerg-13681	241	6	to	to	ADP
fuelectenerg-13681	241	7	changing	change	VERB
fuelectenerg-13681	241	8	conditions	condition	NOUN
fuelectenerg-13681	241	9	.	.	PUNCT
fuelectenerg-13681	242	1	▪	▪	ADJ
fuelectenerg-13681	242	2	robustness	robustness	NOUN
fuelectenerg-13681	242	3	:	:	PUNCT
fuelectenerg-13681	242	4	measures	measure	NOUN
fuelectenerg-13681	242	5	performance	performance	NOUN
fuelectenerg-13681	242	6	degradation	degradation	NOUN
fuelectenerg-13681	242	7	under	under	ADP
fuelectenerg-13681	242	8	channel	channel	NOUN
fuelectenerg-13681	242	9	estimation	estimation	NOUN
fuelectenerg-13681	242	10	errors	error	NOUN
fuelectenerg-13681	242	11	or	or	CCONJ
fuelectenerg-13681	242	12	model	model	NOUN
fuelectenerg-13681	242	13	mismatch	mismatch	NOUN
fuelectenerg-13681	242	14	,	,	PUNCT
fuelectenerg-13681	242	15	defined	define	VERB
fuelectenerg-13681	242	16	as	as	ADP
fuelectenerg-13681	242	17	:	:	PUNCT
fuelectenerg-13681	242	18	perturbed	perturb	VERB
fuelectenerg-13681	242	19	perfect	perfect	ADJ
fuelectenerg-13681	242	20	sinr	sinr	NOUN
fuelectenerg-13681	242	21	robustness	robustness	NOUN
fuelectenerg-13681	242	22	sinr	sinr	NOUN
fuelectenerg-13681	242	23	=	=	SYM
fuelectenerg-13681	242	24	(	(	PUNCT
fuelectenerg-13681	242	25	38	38	NUM
fuelectenerg-13681	242	26	)	)	PUNCT
fuelectenerg-13681	242	27	where	where	SCONJ
fuelectenerg-13681	242	28	sinrperturbed	sinrperturbe	VERB
fuelectenerg-13681	242	29	is	be	AUX
fuelectenerg-13681	242	30	the	the	DET
fuelectenerg-13681	242	31	sinr	sinr	NOUN
fuelectenerg-13681	242	32	achieved	achieve	VERB
fuelectenerg-13681	242	33	with	with	ADP
fuelectenerg-13681	242	34	perturbed	perturb	VERB
fuelectenerg-13681	242	35	channel	channel	NOUN
fuelectenerg-13681	242	36	information	information	NOUN
fuelectenerg-13681	242	37	(	(	PUNCT
fuelectenerg-13681	242	38	containing	contain	VERB
fuelectenerg-13681	242	39	estimation	estimation	NOUN
fuelectenerg-13681	242	40	errors	error	NOUN
fuelectenerg-13681	242	41	)	)	PUNCT
fuelectenerg-13681	242	42	and	and	CCONJ
fuelectenerg-13681	242	43	sinrperfect	sinrperfect	VERB
fuelectenerg-13681	242	44	is	be	AUX
fuelectenerg-13681	242	45	the	the	DET
fuelectenerg-13681	242	46	sinr	sinr	NOUN
fuelectenerg-13681	242	47	with	with	ADP
fuelectenerg-13681	242	48	perfect	perfect	ADJ
fuelectenerg-13681	242	49	channel	channel	NOUN
fuelectenerg-13681	242	50	information	information	NOUN
fuelectenerg-13681	242	51	.	.	PUNCT
fuelectenerg-13681	243	1	values	value	NOUN
fuelectenerg-13681	243	2	closer	close	ADV
fuelectenerg-13681	243	3	to	to	ADP
fuelectenerg-13681	243	4	1	1	NUM
fuelectenerg-13681	243	5	indicate	indicate	VERB
fuelectenerg-13681	243	6	greater	great	ADJ
fuelectenerg-13681	243	7	robustness	robustness	NOUN
fuelectenerg-13681	243	8	.	.	PUNCT
fuelectenerg-13681	244	1	for	for	ADP
fuelectenerg-13681	244	2	each	each	DET
fuelectenerg-13681	244	3	metric	metric	NOUN
fuelectenerg-13681	244	4	,	,	PUNCT
fuelectenerg-13681	244	5	we	we	PRON
fuelectenerg-13681	244	6	compute	compute	VERB
fuelectenerg-13681	244	7	statistical	statistical	ADJ
fuelectenerg-13681	244	8	measures	measure	NOUN
fuelectenerg-13681	244	9	including	include	VERB
fuelectenerg-13681	244	10	mean	mean	NOUN
fuelectenerg-13681	244	11	,	,	PUNCT
fuelectenerg-13681	244	12	variance	variance	NOUN
fuelectenerg-13681	244	13	,	,	PUNCT
fuelectenerg-13681	244	14	95	95	NUM
fuelectenerg-13681	244	15	%	%	NOUN
fuelectenerg-13681	244	16	confidence	confidence	NOUN
fuelectenerg-13681	244	17	intervals	interval	NOUN
fuelectenerg-13681	244	18	,	,	PUNCT
fuelectenerg-13681	244	19	and	and	CCONJ
fuelectenerg-13681	244	20	p	p	X
fuelectenerg-13681	244	21	-	-	PUNCT
fuelectenerg-13681	244	22	values	value	NOUN
fuelectenerg-13681	244	23	to	to	PART
fuelectenerg-13681	244	24	establish	establish	VERB
fuelectenerg-13681	244	25	the	the	DET
fuelectenerg-13681	244	26	statistical	statistical	ADJ
fuelectenerg-13681	244	27	significance	significance	NOUN
fuelectenerg-13681	244	28	of	of	ADP
fuelectenerg-13681	244	29	observed	observed	ADJ
fuelectenerg-13681	244	30	differences	difference	NOUN
fuelectenerg-13681	244	31	.	.	PUNCT
fuelectenerg-13681	245	1	we	we	PRON
fuelectenerg-13681	245	2	use	use	VERB
fuelectenerg-13681	245	3	paired	pair	VERB
fuelectenerg-13681	245	4	t	t	NOUN
fuelectenerg-13681	245	5	-	-	PUNCT
fuelectenerg-13681	245	6	tests	test	NOUN
fuelectenerg-13681	245	7	for	for	ADP
fuelectenerg-13681	245	8	comparing	compare	VERB
fuelectenerg-13681	245	9	algorithm	algorithm	NOUN
fuelectenerg-13681	245	10	performance	performance	NOUN
fuelectenerg-13681	245	11	on	on	ADP
fuelectenerg-13681	245	12	the	the	DET
fuelectenerg-13681	245	13	same	same	ADJ
fuelectenerg-13681	245	14	channel	channel	NOUN
fuelectenerg-13681	245	15	realizations	realization	NOUN
fuelectenerg-13681	245	16	,	,	PUNCT
fuelectenerg-13681	245	17	with	with	ADP
fuelectenerg-13681	245	18	bonferroni	bonferroni	NOUN
fuelectenerg-13681	245	19	correction	correction	NOUN
fuelectenerg-13681	245	20	for	for	ADP
fuelectenerg-13681	245	21	multiple	multiple	ADJ
fuelectenerg-13681	245	22	comparisons	comparison	NOUN
fuelectenerg-13681	245	23	.	.	PUNCT
fuelectenerg-13681	246	1	additionally	additionally	ADV
fuelectenerg-13681	246	2	,	,	PUNCT
fuelectenerg-13681	246	3	we	we	PRON
fuelectenerg-13681	246	4	introduce	introduce	VERB
fuelectenerg-13681	246	5	a	a	DET
fuelectenerg-13681	246	6	composite	composite	ADJ
fuelectenerg-13681	246	7	performance	performance	NOUN
fuelectenerg-13681	246	8	metric	metric	NOUN
fuelectenerg-13681	246	9	that	that	PRON
fuelectenerg-13681	246	10	combines	combine	VERB
fuelectenerg-13681	246	11	multiple	multiple	ADJ
fuelectenerg-13681	246	12	objectives	objective	NOUN
fuelectenerg-13681	246	13	:	:	PUNCT
fuelectenerg-13681	246	14	1	1	NUM
fuelectenerg-13681	246	15	2	2	NUM
fuelectenerg-13681	246	16	3	3	NUM
fuelectenerg-13681	246	17	4	4	NUM
fuelectenerg-13681	246	18	(	(	PUNCT
fuelectenerg-13681	246	19	)	)	PUNCT
fuelectenerg-13681	246	20	sinr	sinr	PROPN
fuelectenerg-13681	246	21	(	(	PUNCT
fuelectenerg-13681	246	22	)	)	PUNCT
fuelectenerg-13681	246	23	ber	ber	PROPN
fuelectenerg-13681	246	24	(	(	PUNCT
fuelectenerg-13681	246	25	)	)	PUNCT
fuelectenerg-13681	246	26	complexity	complexity	NOUN
fuelectenerg-13681	246	27	(	(	PUNCT
fuelectenerg-13681	246	28	)	)	PUNCT
fuelectenerg-13681	246	29	adaptability	adaptability	NOUN
fuelectenerg-13681	246	30	(	(	PUNCT
fuelectenerg-13681	246	31	)	)	PUNCT
fuelectenerg-13681	246	32	j	j	PROPN
fuelectenerg-13681	246	33			NOUN
fuelectenerg-13681	246	34			X
fuelectenerg-13681	246	35			X
fuelectenerg-13681	246	36	=	=	VERB
fuelectenerg-13681	246	37			ADJ
fuelectenerg-13681	246	38	−	−	NOUN
fuelectenerg-13681	246	39			NUM
fuelectenerg-13681	246	40	−	−	NOUN
fuelectenerg-13681	246	41			ADJ
fuelectenerg-13681	246	42	+	+	NUM
fuelectenerg-13681	246	43	w	w	ADJ
fuelectenerg-13681	246	44	w	w	PROPN
fuelectenerg-13681	246	45	w	w	PROPN
fuelectenerg-13681	246	46	w	w	PROPN
fuelectenerg-13681	246	47	w	w	PROPN
fuelectenerg-13681	246	48	(	(	PUNCT
fuelectenerg-13681	246	49	39	39	NUM
fuelectenerg-13681	246	50	)	)	PUNCT
fuelectenerg-13681	246	51	where	where	SCONJ
fuelectenerg-13681	246	52	i	i	PRON
fuelectenerg-13681	246	53	are	be	AUX
fuelectenerg-13681	246	54	weight	weight	NOUN
fuelectenerg-13681	246	55	coefficients	coefficient	NOUN
fuelectenerg-13681	246	56	that	that	PRON
fuelectenerg-13681	246	57	can	can	AUX
fuelectenerg-13681	246	58	be	be	AUX
fuelectenerg-13681	246	59	adjusted	adjust	VERB
fuelectenerg-13681	246	60	based	base	VERB
fuelectenerg-13681	246	61	on	on	ADP
fuelectenerg-13681	246	62	application	application	NOUN
fuelectenerg-13681	246	63	requirements	requirement	NOUN
fuelectenerg-13681	246	64	.	.	PUNCT
fuelectenerg-13681	247	1	3.7	3.7	NUM
fuelectenerg-13681	247	2	.	.	PUNCT
fuelectenerg-13681	248	1	experimental	experimental	ADJ
fuelectenerg-13681	248	2	setup	setup	NOUN
fuelectenerg-13681	248	3	all	all	DET
fuelectenerg-13681	248	4	experiments	experiment	NOUN
fuelectenerg-13681	248	5	were	be	AUX
fuelectenerg-13681	248	6	conducted	conduct	VERB
fuelectenerg-13681	248	7	on	on	ADP
fuelectenerg-13681	248	8	a	a	DET
fuelectenerg-13681	248	9	standardized	standardized	ADJ
fuelectenerg-13681	248	10	hardware	hardware	NOUN
fuelectenerg-13681	248	11	platform	platform	NOUN
fuelectenerg-13681	248	12	to	to	PART
fuelectenerg-13681	248	13	ensure	ensure	VERB
fuelectenerg-13681	248	14	reproducibility	reproducibility	NOUN
fuelectenerg-13681	248	15	:	:	PUNCT
fuelectenerg-13681	248	16	an	an	DET
fuelectenerg-13681	248	17	intel	intel	PROPN
fuelectenerg-13681	248	18	xeon	xeon	PROPN
fuelectenerg-13681	248	19	e5	e5	PROPN
fuelectenerg-13681	248	20	-	-	PUNCT
fuelectenerg-13681	248	21	2680	2680	NUM
fuelectenerg-13681	248	22	v4	v4	NOUN
fuelectenerg-13681	248	23	processor	processor	NOUN
fuelectenerg-13681	248	24	with	with	ADP
fuelectenerg-13681	248	25	128	128	NUM
fuelectenerg-13681	248	26	gb	gb	NOUN
fuelectenerg-13681	248	27	ram	ram	NOUN
fuelectenerg-13681	248	28	and	and	CCONJ
fuelectenerg-13681	248	29	4×	4×	ADJ
fuelectenerg-13681	248	30	nvidia	nvidia	PROPN
fuelectenerg-13681	248	31	tesla	tesla	PROPN
fuelectenerg-13681	248	32	v100	v100	PROPN
fuelectenerg-13681	248	33	gpus	gpus	PROPN
fuelectenerg-13681	248	34	.	.	PUNCT
fuelectenerg-13681	249	1	traditional	traditional	ADJ
fuelectenerg-13681	249	2	optimization	optimization	NOUN
fuelectenerg-13681	249	3	algorithms	algorithm	NOUN
fuelectenerg-13681	249	4	were	be	AUX
fuelectenerg-13681	249	5	implemented	implement	VERB
fuelectenerg-13681	249	6	in	in	ADP
fuelectenerg-13681	249	7	matlab	matlab	PROPN
fuelectenerg-13681	249	8	r2021a	r2021a	PROPN
fuelectenerg-13681	249	9	,	,	PUNCT
fuelectenerg-13681	249	10	while	while	SCONJ
fuelectenerg-13681	249	11	deep	deep	ADJ
fuelectenerg-13681	249	12	learning	learning	NOUN
fuelectenerg-13681	249	13	approaches	approach	NOUN
fuelectenerg-13681	249	14	were	be	AUX
fuelectenerg-13681	249	15	implemented	implement	VERB
fuelectenerg-13681	249	16	in	in	ADP
fuelectenerg-13681	249	17	python	python	NOUN
fuelectenerg-13681	249	18	3.8	3.8	NUM
fuelectenerg-13681	249	19	with	with	ADP
fuelectenerg-13681	249	20	tensorflow	tensorflow	NOUN
fuelectenerg-13681	249	21	2.5	2.5	NUM
fuelectenerg-13681	249	22	.	.	PUNCT
fuelectenerg-13681	250	1	to	to	PART
fuelectenerg-13681	250	2	validate	validate	VERB
fuelectenerg-13681	250	3	performance	performance	NOUN
fuelectenerg-13681	250	4	in	in	ADP
fuelectenerg-13681	250	5	resource	resource	NOUN
fuelectenerg-13681	250	6	-	-	PUNCT
fuelectenerg-13681	250	7	constrained	constrain	VERB
fuelectenerg-13681	250	8	environments	environment	NOUN
fuelectenerg-13681	250	9	,	,	PUNCT
fuelectenerg-13681	250	10	selected	select	VERB
fuelectenerg-13681	250	11	algorithms	algorithm	NOUN
fuelectenerg-13681	250	12	were	be	AUX
fuelectenerg-13681	250	13	also	also	ADV
fuelectenerg-13681	250	14	tested	test	VERB
fuelectenerg-13681	250	15	on	on	ADP
fuelectenerg-13681	250	16	an	an	DET
fuelectenerg-13681	250	17	nvidia	nvidia	PROPN
fuelectenerg-13681	250	18	jetson	jetson	PROPN
fuelectenerg-13681	250	19	agx	agx	PROPN
fuelectenerg-13681	250	20	xavier	xavier	PROPN
fuelectenerg-13681	250	21	embedded	embed	VERB
fuelectenerg-13681	250	22	platform	platform	NOUN
fuelectenerg-13681	250	23	.	.	PUNCT
fuelectenerg-13681	251	1	for	for	ADP
fuelectenerg-13681	251	2	hardware	hardware	NOUN
fuelectenerg-13681	251	3	validation	validation	NOUN
fuelectenerg-13681	251	4	,	,	PUNCT
fuelectenerg-13681	251	5	we	we	PRON
fuelectenerg-13681	251	6	implemented	implement	VERB
fuelectenerg-13681	251	7	key	key	ADJ
fuelectenerg-13681	251	8	algorithms	algorithm	NOUN
fuelectenerg-13681	251	9	on	on	ADP
fuelectenerg-13681	251	10	universal	universal	ADJ
fuelectenerg-13681	251	11	software	software	NOUN
fuelectenerg-13681	251	12	radio	radio	NOUN
fuelectenerg-13681	251	13	peripheral	peripheral	NOUN
fuelectenerg-13681	251	14	(	(	PUNCT
fuelectenerg-13681	251	15	usrp	usrp	NOUN
fuelectenerg-13681	251	16	)	)	PUNCT
fuelectenerg-13681	251	17	x310	x310	NUM
fuelectenerg-13681	251	18	platforms	platform	NOUN
fuelectenerg-13681	251	19	with	with	ADP
fuelectenerg-13681	251	20	ubx-160	ubx-160	PROPN
fuelectenerg-13681	251	21	daughterboards	daughterboard	NOUN
fuelectenerg-13681	251	22	operating	operate	VERB
fuelectenerg-13681	251	23	at	at	ADP
fuelectenerg-13681	251	24	2.4	2.4	NUM
fuelectenerg-13681	251	25	ghz	ghz	NOUN
fuelectenerg-13681	251	26	with	with	ADP
fuelectenerg-13681	251	27	20	20	NUM
fuelectenerg-13681	251	28	mhz	mhz	ADJ
fuelectenerg-13681	251	29	bandwidth	bandwidth	NOUN
fuelectenerg-13681	251	30	.	.	PUNCT
fuelectenerg-13681	252	1	each	each	DET
fuelectenerg-13681	252	2	experiment	experiment	NOUN
fuelectenerg-13681	252	3	was	be	AUX
fuelectenerg-13681	252	4	repeated	repeat	VERB
fuelectenerg-13681	252	5	50	50	NUM
fuelectenerg-13681	252	6	times	time	NOUN
fuelectenerg-13681	252	7	with	with	ADP
fuelectenerg-13681	252	8	different	different	ADJ
fuelectenerg-13681	252	9	random	random	ADJ
fuelectenerg-13681	252	10	seeds	seed	NOUN
fuelectenerg-13681	252	11	to	to	PART
fuelectenerg-13681	252	12	ensure	ensure	VERB
fuelectenerg-13681	252	13	statistical	statistical	ADJ
fuelectenerg-13681	252	14	validity	validity	NOUN
fuelectenerg-13681	252	15	.	.	PUNCT
fuelectenerg-13681	253	1	we	we	PRON
fuelectenerg-13681	253	2	report	report	VERB
fuelectenerg-13681	253	3	95	95	NUM
fuelectenerg-13681	253	4	%	%	NOUN
fuelectenerg-13681	253	5	confidence	confidence	NOUN
fuelectenerg-13681	253	6	intervals	interval	NOUN
fuelectenerg-13681	253	7	for	for	ADP
fuelectenerg-13681	253	8	all	all	DET
fuelectenerg-13681	253	9	results	result	NOUN
fuelectenerg-13681	253	10	and	and	CCONJ
fuelectenerg-13681	253	11	conduct	conduct	VERB
fuelectenerg-13681	253	12	hypothesis	hypothesis	NOUN
fuelectenerg-13681	253	13	testing	testing	NOUN
fuelectenerg-13681	253	14	to	to	PART
fuelectenerg-13681	253	15	establish	establish	VERB
fuelectenerg-13681	253	16	the	the	DET
fuelectenerg-13681	253	17	statistical	statistical	ADJ
fuelectenerg-13681	253	18	significance	significance	NOUN
fuelectenerg-13681	253	19	of	of	ADP
fuelectenerg-13681	253	20	observed	observed	ADJ
fuelectenerg-13681	253	21	differences	difference	NOUN
fuelectenerg-13681	253	22	.	.	PUNCT
fuelectenerg-13681	254	1	for	for	ADP
fuelectenerg-13681	254	2	deep	deep	ADJ
fuelectenerg-13681	254	3	learning	learning	NOUN
fuelectenerg-13681	254	4	approaches	approach	NOUN
fuelectenerg-13681	254	5	,	,	PUNCT
fuelectenerg-13681	254	6	we	we	PRON
fuelectenerg-13681	254	7	used	use	VERB
fuelectenerg-13681	254	8	a	a	DET
fuelectenerg-13681	254	9	training	training	NOUN
fuelectenerg-13681	254	10	/	/	SYM
fuelectenerg-13681	254	11	validation	validation	NOUN
fuelectenerg-13681	254	12	/	/	SYM
fuelectenerg-13681	254	13	test	test	NOUN
fuelectenerg-13681	254	14	split	split	NOUN
fuelectenerg-13681	254	15	of	of	ADP
fuelectenerg-13681	254	16	70%/15%/15	70%/15%/15	NUM
fuelectenerg-13681	254	17	%	%	NOUN
fuelectenerg-13681	254	18	.	.	PUNCT
fuelectenerg-13681	255	1	models	model	NOUN
fuelectenerg-13681	255	2	were	be	AUX
fuelectenerg-13681	255	3	trained	train	VERB
fuelectenerg-13681	255	4	using	use	VERB
fuelectenerg-13681	255	5	early	early	ADV
fuelectenerg-13681	255	6	stopping	stop	VERB
fuelectenerg-13681	255	7	with	with	ADP
fuelectenerg-13681	255	8	a	a	DET
fuelectenerg-13681	255	9	patience	patience	NOUN
fuelectenerg-13681	255	10	of	of	ADP
fuelectenerg-13681	255	11	10	10	NUM
fuelectenerg-13681	255	12	epochs	epoch	NOUN
fuelectenerg-13681	255	13	to	to	PART
fuelectenerg-13681	255	14	prevent	prevent	VERB
fuelectenerg-13681	255	15	overfitting	overfitting	NOUN
fuelectenerg-13681	255	16	.	.	PUNCT
fuelectenerg-13681	256	1	we	we	PRON
fuelectenerg-13681	256	2	also	also	ADV
fuelectenerg-13681	256	3	employed	employ	VERB
fuelectenerg-13681	256	4	k	k	ADJ
fuelectenerg-13681	256	5	-	-	ADJ
fuelectenerg-13681	256	6	fold	fold	ADJ
fuelectenerg-13681	256	7	cross	cross	NOUN
fuelectenerg-13681	256	8	-	-	NOUN
fuelectenerg-13681	256	9	validation	validation	ADJ
fuelectenerg-13681	256	10	(	(	PUNCT
fuelectenerg-13681	256	11	k=5	k=5	X
fuelectenerg-13681	256	12	)	)	PUNCT
fuelectenerg-13681	256	13	to	to	PART
fuelectenerg-13681	256	14	ensure	ensure	VERB
fuelectenerg-13681	256	15	robustness	robustness	NOUN
fuelectenerg-13681	256	16	of	of	ADP
fuelectenerg-13681	256	17	the	the	DET
fuelectenerg-13681	256	18	results	result	NOUN
fuelectenerg-13681	256	19	.	.	PUNCT
fuelectenerg-13681	257	1	4	4	X
fuelectenerg-13681	257	2	.	.	NOUN
fuelectenerg-13681	257	3	results	result	NOUN
fuelectenerg-13681	257	4	and	and	CCONJ
fuelectenerg-13681	257	5	discussion	discussion	NOUN
fuelectenerg-13681	257	6	this	this	DET
fuelectenerg-13681	257	7	section	section	NOUN
fuelectenerg-13681	257	8	presents	present	VERB
fuelectenerg-13681	257	9	our	our	PRON
fuelectenerg-13681	257	10	comparative	comparative	ADJ
fuelectenerg-13681	257	11	performance	performance	NOUN
fuelectenerg-13681	257	12	analysis	analysis	NOUN
fuelectenerg-13681	257	13	between	between	ADP
fuelectenerg-13681	257	14	deep	deep	ADJ
fuelectenerg-13681	257	15	learning	learning	NOUN
fuelectenerg-13681	257	16	approaches	approach	NOUN
fuelectenerg-13681	257	17	and	and	CCONJ
fuelectenerg-13681	257	18	traditional	traditional	ADJ
fuelectenerg-13681	257	19	optimization	optimization	NOUN
fuelectenerg-13681	257	20	algorithms	algorithm	NOUN
fuelectenerg-13681	257	21	for	for	ADP
fuelectenerg-13681	257	22	adaptive	adaptive	ADJ
fuelectenerg-13681	257	23	beamforming	beamforming	NOUN
fuelectenerg-13681	257	24	in	in	ADP
fuelectenerg-13681	257	25	mimo	mimo	PROPN
fuelectenerg-13681	257	26	systems	system	NOUN
fuelectenerg-13681	257	27	.	.	PUNCT
fuelectenerg-13681	258	1	we	we	PRON
fuelectenerg-13681	258	2	analyze	analyze	VERB
fuelectenerg-13681	258	3	the	the	DET
fuelectenerg-13681	258	4	results	result	NOUN
fuelectenerg-13681	258	5	across	across	ADP
fuelectenerg-13681	258	6	various	various	ADJ
fuelectenerg-13681	258	7	scenarios	scenario	NOUN
fuelectenerg-13681	258	8	,	,	PUNCT
fuelectenerg-13681	258	9	highlighting	highlight	VERB
fuelectenerg-13681	258	10	the	the	DET
fuelectenerg-13681	258	11	strengths	strength	NOUN
fuelectenerg-13681	258	12	and	and	CCONJ
fuelectenerg-13681	258	13	limitations	limitation	NOUN
fuelectenerg-13681	258	14	of	of	ADP
fuelectenerg-13681	258	15	each	each	DET
fuelectenerg-13681	258	16	approach	approach	NOUN
fuelectenerg-13681	258	17	.	.	PUNCT
fuelectenerg-13681	259	1	deep	deep	ADJ
fuelectenerg-13681	259	2	learning	learning	NOUN
fuelectenerg-13681	259	3	and	and	CCONJ
fuelectenerg-13681	259	4	traditional	traditional	ADJ
fuelectenerg-13681	259	5	algorithms	algorithm	NOUN
fuelectenerg-13681	259	6	for	for	ADP
fuelectenerg-13681	259	7	mimo	mimo	PROPN
fuelectenerg-13681	259	8	beamforming	beamforme	VERB
fuelectenerg-13681	259	9	209	209	NUM
fuelectenerg-13681	259	10	4.1	4.1	NUM
fuelectenerg-13681	259	11	.	.	PUNCT
fuelectenerg-13681	260	1	performance	performance	NOUN
fuelectenerg-13681	260	2	in	in	ADP
fuelectenerg-13681	260	3	static	static	ADJ
fuelectenerg-13681	260	4	environments	environment	NOUN
fuelectenerg-13681	260	5	fig	fig	NOUN
fuelectenerg-13681	260	6	.	.	PUNCT
fuelectenerg-13681	261	1	1	1	NUM
fuelectenerg-13681	261	2	sinr	sinr	NOUN
fuelectenerg-13681	261	3	performance	performance	NOUN
fuelectenerg-13681	261	4	of	of	ADP
fuelectenerg-13681	261	5	different	different	ADJ
fuelectenerg-13681	261	6	algorithms	algorithm	NOUN
fuelectenerg-13681	261	7	in	in	ADP
fuelectenerg-13681	261	8	static	static	ADJ
fuelectenerg-13681	261	9	channel	channel	NOUN
fuelectenerg-13681	261	10	conditions	condition	NOUN
fuelectenerg-13681	261	11	with	with	ADP
fuelectenerg-13681	261	12	varying	vary	VERB
fuelectenerg-13681	261	13	snr	snr	NOUN
fuelectenerg-13681	261	14	levels	level	NOUN
fuelectenerg-13681	261	15	.	.	PUNCT
fuelectenerg-13681	262	1	deep	deep	ADJ
fuelectenerg-13681	262	2	learning	learning	NOUN
fuelectenerg-13681	262	3	approaches	approach	NOUN
fuelectenerg-13681	262	4	match	match	VERB
fuelectenerg-13681	262	5	traditional	traditional	ADJ
fuelectenerg-13681	262	6	algorithms	algorithm	NOUN
fuelectenerg-13681	262	7	at	at	ADP
fuelectenerg-13681	262	8	medium	medium	NOUN
fuelectenerg-13681	262	9	-	-	PUNCT
fuelectenerg-13681	262	10	to	to	ADP
fuelectenerg-13681	262	11	-	-	PUNCT
fuelectenerg-13681	262	12	high	high	ADJ
fuelectenerg-13681	262	13	snr	snr	NOUN
fuelectenerg-13681	262	14	but	but	CCONJ
fuelectenerg-13681	262	15	underperform	underperform	NOUN
fuelectenerg-13681	262	16	at	at	ADP
fuelectenerg-13681	262	17	low	low	ADJ
fuelectenerg-13681	262	18	snr	snr	PROPN
fuelectenerg-13681	262	19	in	in	ADP
fuelectenerg-13681	262	20	high	high	ADJ
fuelectenerg-13681	262	21	-	-	PUNCT
fuelectenerg-13681	262	22	snr	snr	NOUN
fuelectenerg-13681	262	23	regimes	regime	NOUN
fuelectenerg-13681	262	24	(	(	PUNCT
fuelectenerg-13681	262	25	>	>	SYM
fuelectenerg-13681	262	26	15	15	NUM
fuelectenerg-13681	262	27	db	db	NOUN
fuelectenerg-13681	262	28	)	)	PUNCT
fuelectenerg-13681	262	29	,	,	PUNCT
fuelectenerg-13681	262	30	both	both	PRON
fuelectenerg-13681	262	31	traditional	traditional	ADJ
fuelectenerg-13681	262	32	optimization	optimization	NOUN
fuelectenerg-13681	262	33	algorithms	algorithm	NOUN
fuelectenerg-13681	262	34	and	and	CCONJ
fuelectenerg-13681	262	35	deep	deep	ADJ
fuelectenerg-13681	262	36	learning	learning	NOUN
fuelectenerg-13681	262	37	approaches	approach	NOUN
fuelectenerg-13681	262	38	achieve	achieve	VERB
fuelectenerg-13681	262	39	comparable	comparable	ADJ
fuelectenerg-13681	262	40	performance	performance	NOUN
fuelectenerg-13681	262	41	,	,	PUNCT
fuelectenerg-13681	262	42	with	with	ADP
fuelectenerg-13681	262	43	differences	difference	NOUN
fuelectenerg-13681	262	44	less	less	ADJ
fuelectenerg-13681	262	45	than	than	ADP
fuelectenerg-13681	262	46	0.5	0.5	NUM
fuelectenerg-13681	262	47	db	db	PROPN
fuelectenerg-13681	262	48	(	(	PUNCT
fuelectenerg-13681	262	49	p	p	X
fuelectenerg-13681	262	50	>	>	X
fuelectenerg-13681	262	51	0.05	0.05	NUM
fuelectenerg-13681	262	52	,	,	PUNCT
fuelectenerg-13681	262	53	paired	pair	VERB
fuelectenerg-13681	262	54	t	t	NOUN
fuelectenerg-13681	262	55	-	-	PUNCT
fuelectenerg-13681	262	56	test	test	NOUN
fuelectenerg-13681	262	57	)	)	PUNCT
fuelectenerg-13681	262	58	.	.	PUNCT
fuelectenerg-13681	263	1	socp	socp	PROPN
fuelectenerg-13681	263	2	consistently	consistently	ADV
fuelectenerg-13681	263	3	yields	yield	VERB
fuelectenerg-13681	263	4	the	the	DET
fuelectenerg-13681	263	5	highest	high	ADJ
fuelectenerg-13681	263	6	sinr	sinr	NOUN
fuelectenerg-13681	263	7	(	(	PUNCT
fuelectenerg-13681	263	8	average	average	ADJ
fuelectenerg-13681	263	9	24.8	24.8	NUM
fuelectenerg-13681	263	10	db	db	NOUN
fuelectenerg-13681	263	11	at	at	ADP
fuelectenerg-13681	263	12	20	20	NUM
fuelectenerg-13681	263	13	db	db	PROPN
fuelectenerg-13681	263	14	snr	snr	PROPN
fuelectenerg-13681	263	15	)	)	PUNCT
fuelectenerg-13681	263	16	,	,	PUNCT
fuelectenerg-13681	263	17	followed	follow	VERB
fuelectenerg-13681	263	18	closely	closely	ADV
fuelectenerg-13681	263	19	by	by	ADP
fuelectenerg-13681	263	20	model	model	NOUN
fuelectenerg-13681	263	21	-	-	PUNCT
fuelectenerg-13681	263	22	based	base	VERB
fuelectenerg-13681	263	23	deep	deep	ADJ
fuelectenerg-13681	263	24	learning	learning	NOUN
fuelectenerg-13681	263	25	(	(	PUNCT
fuelectenerg-13681	263	26	24.5	24.5	NUM
fuelectenerg-13681	263	27	db	db	NOUN
fuelectenerg-13681	263	28	)	)	PUNCT
fuelectenerg-13681	263	29	and	and	CCONJ
fuelectenerg-13681	263	30	mvdr	mvdr	NOUN
fuelectenerg-13681	263	31	(	(	PUNCT
fuelectenerg-13681	263	32	24.3	24.3	NUM
fuelectenerg-13681	263	33	db	db	NOUN
fuelectenerg-13681	263	34	)	)	PUNCT
fuelectenerg-13681	263	35	.	.	PUNCT
fuelectenerg-13681	264	1	the	the	DET
fuelectenerg-13681	264	2	performance	performance	NOUN
fuelectenerg-13681	264	3	gap	gap	NOUN
fuelectenerg-13681	264	4	is	be	AUX
fuelectenerg-13681	264	5	not	not	PART
fuelectenerg-13681	264	6	statistically	statistically	ADV
fuelectenerg-13681	264	7	significant	significant	ADJ
fuelectenerg-13681	264	8	in	in	ADP
fuelectenerg-13681	264	9	this	this	DET
fuelectenerg-13681	264	10	regime	regime	NOUN
fuelectenerg-13681	264	11	(	(	PUNCT
fuelectenerg-13681	264	12	p	p	X
fuelectenerg-13681	264	13	=	=	NOUN
fuelectenerg-13681	264	14	0.14	0.14	NUM
fuelectenerg-13681	264	15	)	)	PUNCT
fuelectenerg-13681	264	16	.	.	PUNCT
fuelectenerg-13681	265	1	in	in	ADP
fuelectenerg-13681	265	2	low	low	ADJ
fuelectenerg-13681	265	3	-	-	PUNCT
fuelectenerg-13681	265	4	snr	snr	NOUN
fuelectenerg-13681	265	5	regimes	regime	NOUN
fuelectenerg-13681	265	6	(	(	PUNCT
fuelectenerg-13681	265	7	<	<	X
fuelectenerg-13681	265	8	5	5	NUM
fuelectenerg-13681	265	9	db	db	NOUN
fuelectenerg-13681	265	10	)	)	PUNCT
fuelectenerg-13681	265	11	,	,	PUNCT
fuelectenerg-13681	265	12	traditional	traditional	ADJ
fuelectenerg-13681	265	13	algorithms	algorithm	NOUN
fuelectenerg-13681	265	14	significantly	significantly	ADV
fuelectenerg-13681	265	15	outperform	outperform	VERB
fuelectenerg-13681	265	16	pure	pure	ADJ
fuelectenerg-13681	265	17	deep	deep	ADJ
fuelectenerg-13681	265	18	learning	learning	NOUN
fuelectenerg-13681	265	19	approaches	approach	NOUN
fuelectenerg-13681	265	20	by	by	ADP
fuelectenerg-13681	265	21	1.2–2.8	1.2–2.8	NUM
fuelectenerg-13681	265	22	db	db	X
fuelectenerg-13681	265	23	(	(	PUNCT
fuelectenerg-13681	265	24	p	p	X
fuelectenerg-13681	265	25	<	<	X
fuelectenerg-13681	265	26	0.01	0.01	NUM
fuelectenerg-13681	265	27	)	)	PUNCT
fuelectenerg-13681	265	28	.	.	PUNCT
fuelectenerg-13681	266	1	at	at	ADP
fuelectenerg-13681	266	2	0	0	NUM
fuelectenerg-13681	266	3	db	db	PROPN
fuelectenerg-13681	266	4	snr	snr	PROPN
fuelectenerg-13681	266	5	,	,	PUNCT
fuelectenerg-13681	266	6	socp	socp	PROPN
fuelectenerg-13681	266	7	achieves	achieve	VERB
fuelectenerg-13681	266	8	an	an	DET
fuelectenerg-13681	266	9	average	average	ADJ
fuelectenerg-13681	266	10	sinr	sinr	NOUN
fuelectenerg-13681	266	11	of	of	ADP
fuelectenerg-13681	266	12	8.7	8.7	NUM
fuelectenerg-13681	266	13	db	db	NOUN
fuelectenerg-13681	266	14	,	,	PUNCT
fuelectenerg-13681	266	15	compared	compare	VERB
fuelectenerg-13681	266	16	to	to	ADP
fuelectenerg-13681	266	17	6.5	6.5	NUM
fuelectenerg-13681	266	18	db	db	NOUN
fuelectenerg-13681	266	19	for	for	ADP
fuelectenerg-13681	266	20	the	the	DET
fuelectenerg-13681	266	21	best	well	ADV
fuelectenerg-13681	266	22	-	-	PUNCT
fuelectenerg-13681	266	23	performing	perform	VERB
fuelectenerg-13681	266	24	deep	deep	ADJ
fuelectenerg-13681	266	25	learning	learning	NOUN
fuelectenerg-13681	266	26	method	method	NOUN
fuelectenerg-13681	266	27	(	(	PUNCT
fuelectenerg-13681	266	28	mbdl	mbdl	PROPN
fuelectenerg-13681	266	29	)	)	PUNCT
fuelectenerg-13681	266	30	.	.	PUNCT
fuelectenerg-13681	267	1	this	this	DET
fuelectenerg-13681	267	2	performance	performance	NOUN
fuelectenerg-13681	267	3	gap	gap	NOUN
fuelectenerg-13681	267	4	can	can	AUX
fuelectenerg-13681	267	5	be	be	AUX
fuelectenerg-13681	267	6	attributed	attribute	VERB
fuelectenerg-13681	267	7	to	to	ADP
fuelectenerg-13681	267	8	the	the	DET
fuelectenerg-13681	267	9	fact	fact	NOUN
fuelectenerg-13681	267	10	that	that	SCONJ
fuelectenerg-13681	267	11	traditional	traditional	ADJ
fuelectenerg-13681	267	12	algorithms	algorithm	NOUN
fuelectenerg-13681	267	13	incorporate	incorporate	VERB
fuelectenerg-13681	267	14	optimal	optimal	ADJ
fuelectenerg-13681	267	15	statistical	statistical	ADJ
fuelectenerg-13681	267	16	estimators	estimator	NOUN
fuelectenerg-13681	267	17	specifically	specifically	ADV
fuelectenerg-13681	267	18	designed	design	VERB
fuelectenerg-13681	267	19	for	for	ADP
fuelectenerg-13681	267	20	low	low	ADJ
fuelectenerg-13681	267	21	-	-	PUNCT
fuelectenerg-13681	267	22	snr	snr	NOUN
fuelectenerg-13681	267	23	conditions	condition	NOUN
fuelectenerg-13681	267	24	,	,	PUNCT
fuelectenerg-13681	267	25	while	while	SCONJ
fuelectenerg-13681	267	26	neural	neural	ADJ
fuelectenerg-13681	267	27	networks	network	NOUN
fuelectenerg-13681	267	28	struggle	struggle	VERB
fuelectenerg-13681	267	29	to	to	PART
fuelectenerg-13681	267	30	generalize	generalize	VERB
fuelectenerg-13681	267	31	to	to	ADP
fuelectenerg-13681	267	32	these	these	DET
fuelectenerg-13681	267	33	more	more	ADV
fuelectenerg-13681	267	34	challenging	challenging	ADJ
fuelectenerg-13681	267	35	scenarios	scenario	NOUN
fuelectenerg-13681	267	36	.	.	PUNCT
fuelectenerg-13681	268	1	theoretical	theoretical	ADJ
fuelectenerg-13681	268	2	analysis	analysis	NOUN
fuelectenerg-13681	268	3	confirms	confirm	VERB
fuelectenerg-13681	268	4	these	these	DET
fuelectenerg-13681	268	5	empirical	empirical	ADJ
fuelectenerg-13681	268	6	observations	observation	NOUN
fuelectenerg-13681	268	7	.	.	PUNCT
fuelectenerg-13681	269	1	for	for	ADP
fuelectenerg-13681	269	2	mvdr	mvdr	NOUN
fuelectenerg-13681	269	3	beamforming	beamforming	NOUN
fuelectenerg-13681	269	4	,	,	PUNCT
fuelectenerg-13681	269	5	the	the	DET
fuelectenerg-13681	269	6	sinr	sinr	NOUN
fuelectenerg-13681	269	7	is	be	AUX
fuelectenerg-13681	269	8	bounded	bound	VERB
fuelectenerg-13681	269	9	by	by	ADP
fuelectenerg-13681	269	10	:	:	PUNCT
fuelectenerg-13681	269	11	2	2	NUM
fuelectenerg-13681	269	12	1	1	NUM
fuelectenerg-13681	269	13	min	min	NOUN
fuelectenerg-13681	269	14	2	2	NUM
fuelectenerg-13681	269	15	1	1	NUM
fuelectenerg-13681	269	16	s	s	PART
fuelectenerg-13681	269	17	(	(	PUNCT
fuelectenerg-13681	269	18	inr	inr	PROPN
fuelectenerg-13681	269	19	tr	tr	VERB
fuelectenerg-13681	269	20	(	(	PUNCT
fuelectenerg-13681	269	21	)	)	PUNCT
fuelectenerg-13681	269	22	)	)	PUNCT
fuelectenerg-13681	270	1	s	s	PART
fuelectenerg-13681	270	2	s	s	X
fuelectenerg-13681	270	3	n	n	NOUN
fuelectenerg-13681	270	4	mvdr	mvdr	NOUN
fuelectenerg-13681	270	5	n	n	PROPN
fuelectenerg-13681	270	6	s	s	VERB
fuelectenerg-13681	270	7	n	n	PRON
fuelectenerg-13681	270	8			PROPN
fuelectenerg-13681	270	9			X
fuelectenerg-13681	270	10			X
fuelectenerg-13681	270	11	−	−	NOUN
fuelectenerg-13681	270	12	−	−	NOUN
fuelectenerg-13681	270	13			NUM
fuelectenerg-13681	271	1			NUM
fuelectenerg-13681	271	2	r	r	NOUN
fuelectenerg-13681	271	3	r	r	NOUN
fuelectenerg-13681	271	4	r	r	NOUN
fuelectenerg-13681	271	5	r	r	NOUN
fuelectenerg-13681	271	6	(	(	PUNCT
fuelectenerg-13681	271	7	40	40	NUM
fuelectenerg-13681	271	8	)	)	PUNCT
fuelectenerg-13681	271	9	where	where	SCONJ
fuelectenerg-13681	271	10	rs	rs	NOUN
fuelectenerg-13681	271	11	and	and	CCONJ
fuelectenerg-13681	271	12	rn	rn	PROPN
fuelectenerg-13681	271	13	are	be	AUX
fuelectenerg-13681	271	14	the	the	DET
fuelectenerg-13681	271	15	signal	signal	ADJ
fuelectenerg-13681	271	16	and	and	CCONJ
fuelectenerg-13681	271	17	noise	noise	NOUN
fuelectenerg-13681	271	18	correlation	correlation	NOUN
fuelectenerg-13681	271	19	matrices	matrix	NOUN
fuelectenerg-13681	271	20	,	,	PUNCT
fuelectenerg-13681	271	21	respectively	respectively	ADV
fuelectenerg-13681	271	22	.	.	PUNCT
fuelectenerg-13681	272	1	this	this	DET
fuelectenerg-13681	272	2	bound	bind	VERB
fuelectenerg-13681	272	3	becomes	become	VERB
fuelectenerg-13681	272	4	tighter	tight	ADJ
fuelectenerg-13681	272	5	as	as	SCONJ
fuelectenerg-13681	272	6	snr	snr	PROPN
fuelectenerg-13681	272	7	decreases	decrease	VERB
fuelectenerg-13681	272	8	,	,	PUNCT
fuelectenerg-13681	272	9	explaining	explain	VERB
fuelectenerg-13681	272	10	the	the	DET
fuelectenerg-13681	272	11	superior	superior	ADJ
fuelectenerg-13681	272	12	performance	performance	NOUN
fuelectenerg-13681	272	13	of	of	ADP
fuelectenerg-13681	272	14	traditional	traditional	ADJ
fuelectenerg-13681	272	15	algorithms	algorithm	NOUN
fuelectenerg-13681	272	16	in	in	ADP
fuelectenerg-13681	272	17	low	low	ADJ
fuelectenerg-13681	272	18	-	-	PUNCT
fuelectenerg-13681	272	19	snr	snr	NOUN
fuelectenerg-13681	272	20	conditions	condition	NOUN
fuelectenerg-13681	272	21	.	.	PUNCT
fuelectenerg-13681	273	1	table	table	NOUN
fuelectenerg-13681	273	2	2	2	NUM
fuelectenerg-13681	273	3	presents	present	NOUN
fuelectenerg-13681	273	4	the	the	DET
fuelectenerg-13681	273	5	bit	bit	NOUN
fuelectenerg-13681	273	6	error	error	NOUN
fuelectenerg-13681	273	7	rate	rate	NOUN
fuelectenerg-13681	273	8	performance	performance	NOUN
fuelectenerg-13681	273	9	across	across	ADP
fuelectenerg-13681	273	10	different	different	ADJ
fuelectenerg-13681	273	11	modulation	modulation	NOUN
fuelectenerg-13681	273	12	schemes	scheme	NOUN
fuelectenerg-13681	273	13	for	for	ADP
fuelectenerg-13681	273	14	all	all	DET
fuelectenerg-13681	273	15	evaluated	evaluated	ADJ
fuelectenerg-13681	273	16	algorithms	algorithm	NOUN
fuelectenerg-13681	273	17	.	.	PUNCT
fuelectenerg-13681	274	1	210	210	NUM
fuelectenerg-13681	274	2	m.	m.	NOUN
fuelectenerg-13681	274	3	mohamed	mohamed	PROPN
fuelectenerg-13681	274	4	,	,	PUNCT
fuelectenerg-13681	274	5	f.	f.	PROPN
fuelectenerg-13681	274	6	aljuaid	aljuaid	PROPN
fuelectenerg-13681	274	7	,	,	PUNCT
fuelectenerg-13681	274	8	mhd	mhd	PROPN
fuelectenerg-13681	274	9	w.	w.	PROPN
fuelectenerg-13681	274	10	koubeisi	koubeisi	PROPN
fuelectenerg-13681	274	11	table	table	NOUN
fuelectenerg-13681	274	12	2	2	NUM
fuelectenerg-13681	274	13	bit	bit	NOUN
fuelectenerg-13681	274	14	error	error	NOUN
fuelectenerg-13681	274	15	rate	rate	NOUN
fuelectenerg-13681	274	16	(	(	PUNCT
fuelectenerg-13681	274	17	ber	ber	NOUN
fuelectenerg-13681	274	18	)	)	PUNCT
fuelectenerg-13681	274	19	performance	performance	NOUN
fuelectenerg-13681	274	20	in	in	ADP
fuelectenerg-13681	274	21	static	static	ADJ
fuelectenerg-13681	274	22	channels	channel	NOUN
fuelectenerg-13681	274	23	2*algorithm	2*algorithm	NUM
fuelectenerg-13681	274	24	qpsk	qpsk	NOUN
fuelectenerg-13681	274	25	16	16	NUM
fuelectenerg-13681	274	26	-	-	PUNCT
fuelectenerg-13681	274	27	qam	qam	NOUN
fuelectenerg-13681	274	28	64	64	NUM
fuelectenerg-13681	274	29	-	-	PUNCT
fuelectenerg-13681	274	30	qam	qam	PROPN
fuelectenerg-13681	274	31	lms	lm	NOUN
fuelectenerg-13681	274	32	1.82	1.82	NUM
fuelectenerg-13681	274	33			PROPN
fuelectenerg-13681	274	34	.11	.11	NUM
fuelectenerg-13681	274	35	5.64	5.64	NUM
fuelectenerg-13681	274	36			PROPN
fuelectenerg-13681	274	37	.24	.24	NUM
fuelectenerg-13681	274	38	18.92	18.92	NUM
fuelectenerg-13681	275	1			PROPN
fuelectenerg-13681	275	2	.73	.73	NUM
fuelectenerg-13681	275	3	rls	rls	PROPN
fuelectenerg-13681	275	4	1.43	1.43	NUM
fuelectenerg-13681	275	5			PROPN
fuelectenerg-13681	275	6	.09	.09	NUM
fuelectenerg-13681	275	7	4.21	4.21	NUM
fuelectenerg-13681	275	8			PROPN
fuelectenerg-13681	275	9	.18	.18	NUM
fuelectenerg-13681	275	10	15.37	15.37	NUM
fuelectenerg-13681	275	11			PROPN
fuelectenerg-13681	275	12	.65	.65	NUM
fuelectenerg-13681	275	13	mvdr	mvdr	VERB
fuelectenerg-13681	275	14	1.25	1.25	NUM
fuelectenerg-13681	275	15			PROPN
fuelectenerg-13681	275	16	.07	.07	NUM
fuelectenerg-13681	275	17	3.86	3.86	NUM
fuelectenerg-13681	275	18			PROPN
fuelectenerg-13681	275	19	.15	.15	NUM
fuelectenerg-13681	275	20	14.52	14.52	NUM
fuelectenerg-13681	275	21			PROPN
fuelectenerg-13681	275	22	.59	.59	NUM
fuelectenerg-13681	275	23	socp	socp	NOUN
fuelectenerg-13681	275	24	1.18	1.18	NUM
fuelectenerg-13681	275	25			PROPN
fuelectenerg-13681	275	26	.06	.06	NUM
fuelectenerg-13681	275	27	3.54	3.54	NUM
fuelectenerg-13681	275	28			PROPN
fuelectenerg-13681	275	29	.14	.14	NUM
fuelectenerg-13681	275	30	13.68	13.68	NUM
fuelectenerg-13681	275	31			PROPN
fuelectenerg-13681	275	32	.54	.54	NUM
fuelectenerg-13681	275	33	fc	fc	PROPN
fuelectenerg-13681	275	34	-	-	PUNCT
fuelectenerg-13681	275	35	dnn	dnn	PROPN
fuelectenerg-13681	275	36	1.95	1.95	NUM
fuelectenerg-13681	275	37			PROPN
fuelectenerg-13681	276	1	.12	.12	NUM
fuelectenerg-13681	276	2	6.12	6.12	NUM
fuelectenerg-13681	276	3			PROPN
fuelectenerg-13681	276	4	.27	.27	NUM
fuelectenerg-13681	276	5	21.43	21.43	NUM
fuelectenerg-13681	277	1			PROPN
fuelectenerg-13681	277	2	.85	.85	NUM
fuelectenerg-13681	277	3	cnn	cnn	NOUN
fuelectenerg-13681	277	4	1.53	1.53	NUM
fuelectenerg-13681	277	5			PROPN
fuelectenerg-13681	277	6	.10	.10	NUM
fuelectenerg-13681	277	7	4.86	4.86	NUM
fuelectenerg-13681	277	8			PROPN
fuelectenerg-13681	277	9	.21	.21	NUM
fuelectenerg-13681	277	10	17.25	17.25	NUM
fuelectenerg-13681	277	11			PROPN
fuelectenerg-13681	277	12	.73	.73	NUM
fuelectenerg-13681	277	13	rnn	rnn	PROPN
fuelectenerg-13681	277	14	-	-	PUNCT
fuelectenerg-13681	277	15	lstm	lstm	ADJ
fuelectenerg-13681	277	16	1.78	1.78	NUM
fuelectenerg-13681	277	17			PROPN
fuelectenerg-13681	277	18	.11	.11	NUM
fuelectenerg-13681	277	19	5.42	5.42	NUM
fuelectenerg-13681	277	20			PROPN
fuelectenerg-13681	277	21	.25	.25	NUM
fuelectenerg-13681	277	22	19.67	19.67	NUM
fuelectenerg-13681	277	23			PROPN
fuelectenerg-13681	277	24	.78	.78	NUM
fuelectenerg-13681	277	25	mbdl	mbdl	NOUN
fuelectenerg-13681	277	26	1.35	1.35	NUM
fuelectenerg-13681	277	27			PROPN
fuelectenerg-13681	277	28	.08	.08	NUM
fuelectenerg-13681	277	29	4.18	4.18	NUM
fuelectenerg-13681	277	30			PROPN
fuelectenerg-13681	277	31	.17	.17	NUM
fuelectenerg-13681	277	32	16.32	16.32	NUM
fuelectenerg-13681	277	33			PROPN
fuelectenerg-13681	277	34	.68	.68	NUM
fuelectenerg-13681	277	35	for	for	ADP
fuelectenerg-13681	277	36	qpsk	qpsk	NOUN
fuelectenerg-13681	277	37	modulation	modulation	NOUN
fuelectenerg-13681	277	38	at	at	ADP
fuelectenerg-13681	277	39	high	high	ADJ
fuelectenerg-13681	277	40	snr	snr	PROPN
fuelectenerg-13681	277	41	(	(	PUNCT
fuelectenerg-13681	277	42	20	20	NUM
fuelectenerg-13681	277	43	db	db	NOUN
fuelectenerg-13681	277	44	)	)	PUNCT
fuelectenerg-13681	277	45	,	,	PUNCT
fuelectenerg-13681	277	46	all	all	DET
fuelectenerg-13681	277	47	algorithms	algorithm	NOUN
fuelectenerg-13681	277	48	achieve	achieve	VERB
fuelectenerg-13681	277	49	similar	similar	ADJ
fuelectenerg-13681	277	50	ber	ber	NOUN
fuelectenerg-13681	277	51	(	(	PUNCT
fuelectenerg-13681	277	52	differences	difference	NOUN
fuelectenerg-13681	277	53	<	<	X
fuelectenerg-13681	277	54	5	5	NUM
fuelectenerg-13681	277	55			NOUN
fuelectenerg-13681	277	56	10−4p	10−4p	NUM
fuelectenerg-13681	277	57	>	>	X
fuelectenerg-13681	277	58	0.05	0.05	NUM
fuelectenerg-13681	277	59	)	)	PUNCT
fuelectenerg-13681	277	60	.	.	PUNCT
fuelectenerg-13681	278	1	for	for	ADP
fuelectenerg-13681	278	2	higher	high	ADJ
fuelectenerg-13681	278	3	-	-	PUNCT
fuelectenerg-13681	278	4	order	order	NOUN
fuelectenerg-13681	278	5	modulations	modulation	NOUN
fuelectenerg-13681	278	6	(	(	PUNCT
fuelectenerg-13681	278	7	16	16	NUM
fuelectenerg-13681	278	8	-	-	PUNCT
fuelectenerg-13681	278	9	qam	qam	NOUN
fuelectenerg-13681	278	10	,	,	PUNCT
fuelectenerg-13681	278	11	64	64	NUM
fuelectenerg-13681	278	12	-	-	PUNCT
fuelectenerg-13681	278	13	qam	qam	NOUN
fuelectenerg-13681	278	14	)	)	PUNCT
fuelectenerg-13681	278	15	,	,	PUNCT
fuelectenerg-13681	278	16	traditional	traditional	ADJ
fuelectenerg-13681	278	17	algorithms	algorithm	NOUN
fuelectenerg-13681	278	18	demonstrate	demonstrate	VERB
fuelectenerg-13681	278	19	a	a	DET
fuelectenerg-13681	278	20	statistically	statistically	ADV
fuelectenerg-13681	278	21	significant	significant	ADJ
fuelectenerg-13681	278	22	performance	performance	NOUN
fuelectenerg-13681	278	23	advantage	advantage	NOUN
fuelectenerg-13681	278	24	(	(	PUNCT
fuelectenerg-13681	278	25	p	p	X
fuelectenerg-13681	278	26	<	<	X
fuelectenerg-13681	278	27	0.01	0.01	NUM
fuelectenerg-13681	278	28	)	)	PUNCT
fuelectenerg-13681	278	29	,	,	PUNCT
fuelectenerg-13681	278	30	especially	especially	ADV
fuelectenerg-13681	278	31	in	in	ADP
fuelectenerg-13681	278	32	challenging	challenging	ADJ
fuelectenerg-13681	278	33	channel	channel	NOUN
fuelectenerg-13681	278	34	conditions	condition	NOUN
fuelectenerg-13681	278	35	.	.	PUNCT
fuelectenerg-13681	279	1	the	the	DET
fuelectenerg-13681	279	2	cnn	cnn	PROPN
fuelectenerg-13681	279	3	architecture	architecture	NOUN
fuelectenerg-13681	279	4	achieves	achieve	VERB
fuelectenerg-13681	279	5	ber	ber	NOUN
fuelectenerg-13681	279	6	within	within	ADP
fuelectenerg-13681	279	7	15	15	NUM
fuelectenerg-13681	279	8	%	%	NOUN
fuelectenerg-13681	279	9	of	of	ADP
fuelectenerg-13681	279	10	traditional	traditional	ADJ
fuelectenerg-13681	279	11	methods	method	NOUN
fuelectenerg-13681	279	12	for	for	ADP
fuelectenerg-13681	279	13	16	16	NUM
fuelectenerg-13681	279	14	-	-	PUNCT
fuelectenerg-13681	279	15	qam	qam	PROPN
fuelectenerg-13681	279	16	,	,	PUNCT
fuelectenerg-13681	279	17	making	make	VERB
fuelectenerg-13681	279	18	it	it	PRON
fuelectenerg-13681	279	19	the	the	DET
fuelectenerg-13681	279	20	best	well	ADV
fuelectenerg-13681	279	21	-	-	PUNCT
fuelectenerg-13681	279	22	performing	perform	VERB
fuelectenerg-13681	279	23	pure	pure	ADJ
fuelectenerg-13681	279	24	deep	deep	ADJ
fuelectenerg-13681	279	25	learning	learning	NOUN
fuelectenerg-13681	279	26	approach	approach	NOUN
fuelectenerg-13681	279	27	for	for	ADP
fuelectenerg-13681	279	28	this	this	DET
fuelectenerg-13681	279	29	metric	metric	NOUN
fuelectenerg-13681	279	30	.	.	PUNCT
fuelectenerg-13681	280	1	the	the	DET
fuelectenerg-13681	280	2	model	model	NOUN
fuelectenerg-13681	280	3	-	-	PUNCT
fuelectenerg-13681	280	4	based	base	VERB
fuelectenerg-13681	280	5	deep	deep	ADJ
fuelectenerg-13681	280	6	learning	learning	NOUN
fuelectenerg-13681	280	7	(	(	PUNCT
fuelectenerg-13681	280	8	mbdl	mbdl	NOUN
fuelectenerg-13681	280	9	)	)	PUNCT
fuelectenerg-13681	280	10	approach	approach	NOUN
fuelectenerg-13681	280	11	performs	perform	VERB
fuelectenerg-13681	280	12	significantly	significantly	ADV
fuelectenerg-13681	280	13	better	well	ADJ
fuelectenerg-13681	280	14	than	than	ADP
fuelectenerg-13681	280	15	other	other	ADJ
fuelectenerg-13681	280	16	neural	neural	ADJ
fuelectenerg-13681	280	17	architectures	architecture	NOUN
fuelectenerg-13681	280	18	(	(	PUNCT
fuelectenerg-13681	280	19	p	p	X
fuelectenerg-13681	280	20	<	<	X
fuelectenerg-13681	280	21	0.01	0.01	NUM
fuelectenerg-13681	280	22	)	)	PUNCT
fuelectenerg-13681	280	23	,	,	PUNCT
fuelectenerg-13681	280	24	achieving	achieve	VERB
fuelectenerg-13681	280	25	ber	ber	NOUN
fuelectenerg-13681	280	26	within	within	ADP
fuelectenerg-13681	280	27	8	8	NUM
fuelectenerg-13681	280	28	%	%	NOUN
fuelectenerg-13681	280	29	of	of	ADP
fuelectenerg-13681	280	30	socp	socp	NOUN
fuelectenerg-13681	280	31	for	for	ADP
fuelectenerg-13681	280	32	all	all	DET
fuelectenerg-13681	280	33	modulation	modulation	NOUN
fuelectenerg-13681	280	34	schemes	scheme	NOUN
fuelectenerg-13681	280	35	.	.	PUNCT
fuelectenerg-13681	281	1	this	this	PRON
fuelectenerg-13681	281	2	highlights	highlight	VERB
fuelectenerg-13681	281	3	the	the	DET
fuelectenerg-13681	281	4	advantage	advantage	NOUN
fuelectenerg-13681	281	5	of	of	ADP
fuelectenerg-13681	281	6	incorporating	incorporate	VERB
fuelectenerg-13681	281	7	domain	domain	NOUN
fuelectenerg-13681	281	8	knowledge	knowledge	NOUN
fuelectenerg-13681	281	9	into	into	ADP
fuelectenerg-13681	281	10	neural	neural	ADJ
fuelectenerg-13681	281	11	network	network	NOUN
fuelectenerg-13681	281	12	design	design	NOUN
fuelectenerg-13681	281	13	.	.	PUNCT
fuelectenerg-13681	282	1	statistical	statistical	ADJ
fuelectenerg-13681	282	2	analysis	analysis	NOUN
fuelectenerg-13681	282	3	of	of	ADP
fuelectenerg-13681	282	4	the	the	DET
fuelectenerg-13681	282	5	ber	ber	NOUN
fuelectenerg-13681	282	6	results	result	NOUN
fuelectenerg-13681	282	7	shows	show	VERB
fuelectenerg-13681	282	8	a	a	DET
fuelectenerg-13681	282	9	strong	strong	ADJ
fuelectenerg-13681	282	10	correlation	correlation	NOUN
fuelectenerg-13681	282	11	(	(	PUNCT
fuelectenerg-13681	282	12	r	r	NOUN
fuelectenerg-13681	282	13	=	=	SYM
fuelectenerg-13681	282	14	0.92	0.92	NUM
fuelectenerg-13681	282	15	,	,	PUNCT
fuelectenerg-13681	282	16	p	p	X
fuelectenerg-13681	282	17	<	<	X
fuelectenerg-13681	282	18	0.001	0.001	NUM
fuelectenerg-13681	282	19	)	)	PUNCT
fuelectenerg-13681	282	20	between	between	ADP
fuelectenerg-13681	282	21	algorithm	algorithm	NOUN
fuelectenerg-13681	282	22	performance	performance	NOUN
fuelectenerg-13681	282	23	and	and	CCONJ
fuelectenerg-13681	282	24	channel	channel	NOUN
fuelectenerg-13681	282	25	condition	condition	NOUN
fuelectenerg-13681	282	26	number	number	NOUN
fuelectenerg-13681	282	27	,	,	PUNCT
fuelectenerg-13681	282	28	with	with	ADP
fuelectenerg-13681	282	29	all	all	DET
fuelectenerg-13681	282	30	algorithms	algorithm	NOUN
fuelectenerg-13681	282	31	exhibiting	exhibit	VERB
fuelectenerg-13681	282	32	degraded	degraded	ADJ
fuelectenerg-13681	282	33	performance	performance	NOUN
fuelectenerg-13681	282	34	in	in	ADP
fuelectenerg-13681	282	35	ill	ill	ADV
fuelectenerg-13681	282	36	-	-	PUNCT
fuelectenerg-13681	282	37	conditioned	condition	VERB
fuelectenerg-13681	282	38	channels	channel	NOUN
fuelectenerg-13681	282	39	(	(	PUNCT
fuelectenerg-13681	282	40	high	high	ADJ
fuelectenerg-13681	282	41	condition	condition	NOUN
fuelectenerg-13681	282	42	number	number	NOUN
fuelectenerg-13681	282	43	)	)	PUNCT
fuelectenerg-13681	282	44	.	.	PUNCT
fuelectenerg-13681	283	1	traditional	traditional	ADJ
fuelectenerg-13681	283	2	algorithms	algorithm	NOUN
fuelectenerg-13681	283	3	demonstrate	demonstrate	VERB
fuelectenerg-13681	283	4	greater	great	ADJ
fuelectenerg-13681	283	5	robustness	robustness	NOUN
fuelectenerg-13681	283	6	to	to	ADP
fuelectenerg-13681	283	7	ill	ill	ADJ
fuelectenerg-13681	283	8	-	-	PUNCT
fuelectenerg-13681	283	9	conditioning	conditioning	NOUN
fuelectenerg-13681	283	10	,	,	PUNCT
fuelectenerg-13681	283	11	with	with	ADP
fuelectenerg-13681	283	12	performance	performance	NOUN
fuelectenerg-13681	283	13	degradation	degradation	NOUN
fuelectenerg-13681	283	14	of	of	ADP
fuelectenerg-13681	283	15	28	28	NUM
fuelectenerg-13681	283	16	%	%	NOUN
fuelectenerg-13681	283	17	compared	compare	VERB
fuelectenerg-13681	283	18	to	to	ADP
fuelectenerg-13681	283	19	41	41	NUM
fuelectenerg-13681	283	20	%	%	NOUN
fuelectenerg-13681	283	21	for	for	ADP
fuelectenerg-13681	283	22	pure	pure	ADJ
fuelectenerg-13681	283	23	deep	deep	ADJ
fuelectenerg-13681	283	24	learning	learning	NOUN
fuelectenerg-13681	283	25	approaches	approach	NOUN
fuelectenerg-13681	283	26	.	.	PUNCT
fuelectenerg-13681	284	1	4.2	4.2	NUM
fuelectenerg-13681	284	2	.	.	PUNCT
fuelectenerg-13681	284	3	performance	performance	NOUN
fuelectenerg-13681	284	4	in	in	ADP
fuelectenerg-13681	284	5	dynamic	dynamic	ADJ
fuelectenerg-13681	284	6	environments	environment	NOUN
fuelectenerg-13681	284	7	fig	fig	NOUN
fuelectenerg-13681	284	8	.	.	PUNCT
fuelectenerg-13681	284	9	2	2	NUM
fuelectenerg-13681	284	10	sinr	sinr	NOUN
fuelectenerg-13681	284	11	tracking	tracking	NOUN
fuelectenerg-13681	284	12	performance	performance	NOUN
fuelectenerg-13681	284	13	in	in	ADP
fuelectenerg-13681	284	14	dynamic	dynamic	ADJ
fuelectenerg-13681	284	15	channel	channel	NOUN
fuelectenerg-13681	284	16	conditions	condition	NOUN
fuelectenerg-13681	284	17	with	with	ADP
fuelectenerg-13681	284	18	varying	vary	VERB
fuelectenerg-13681	284	19	user	user	NOUN
fuelectenerg-13681	284	20	mobility	mobility	NOUN
fuelectenerg-13681	284	21	.	.	PUNCT
fuelectenerg-13681	285	1	deep	deep	ADJ
fuelectenerg-13681	285	2	learning	learning	NOUN
fuelectenerg-13681	285	3	approaches	approach	NOUN
fuelectenerg-13681	285	4	demonstrate	demonstrate	VERB
fuelectenerg-13681	285	5	superior	superior	ADJ
fuelectenerg-13681	285	6	adaptation	adaptation	NOUN
fuelectenerg-13681	285	7	capabilities	capability	NOUN
fuelectenerg-13681	285	8	in	in	ADP
fuelectenerg-13681	285	9	high	high	ADJ
fuelectenerg-13681	285	10	-	-	PUNCT
fuelectenerg-13681	285	11	mobility	mobility	NOUN
fuelectenerg-13681	285	12	scenarios	scenario	NOUN
fuelectenerg-13681	285	13	deep	deep	ADJ
fuelectenerg-13681	285	14	learning	learning	NOUN
fuelectenerg-13681	285	15	and	and	CCONJ
fuelectenerg-13681	285	16	traditional	traditional	ADJ
fuelectenerg-13681	285	17	algorithms	algorithm	NOUN
fuelectenerg-13681	285	18	for	for	ADP
fuelectenerg-13681	285	19	mimo	mimo	PROPN
fuelectenerg-13681	285	20	beamforming	beamforme	VERB
fuelectenerg-13681	285	21	211	211	NUM
fuelectenerg-13681	285	22	in	in	ADP
fuelectenerg-13681	285	23	low	low	ADJ
fuelectenerg-13681	285	24	-	-	PUNCT
fuelectenerg-13681	285	25	mobility	mobility	NOUN
fuelectenerg-13681	285	26	scenarios	scenario	NOUN
fuelectenerg-13681	285	27	(	(	PUNCT
fuelectenerg-13681	285	28	pedestrian	pedestrian	NOUN
fuelectenerg-13681	285	29	speed	speed	NOUN
fuelectenerg-13681	285	30	,	,	PUNCT
fuelectenerg-13681	285	31	3	3	NUM
fuelectenerg-13681	285	32	km	km	NOUN
fuelectenerg-13681	285	33	/	/	SYM
fuelectenerg-13681	285	34	h	h	NOUN
fuelectenerg-13681	285	35	)	)	PUNCT
fuelectenerg-13681	285	36	,	,	PUNCT
fuelectenerg-13681	285	37	both	both	CCONJ
fuelectenerg-13681	285	38	traditional	traditional	ADJ
fuelectenerg-13681	285	39	and	and	CCONJ
fuelectenerg-13681	285	40	deep	deep	ADJ
fuelectenerg-13681	285	41	learning	learning	NOUN
fuelectenerg-13681	285	42	approaches	approach	NOUN
fuelectenerg-13681	285	43	show	show	VERB
fuelectenerg-13681	285	44	similar	similar	ADJ
fuelectenerg-13681	285	45	adaptation	adaptation	NOUN
fuelectenerg-13681	285	46	capabilities	capability	NOUN
fuelectenerg-13681	285	47	.	.	PUNCT
fuelectenerg-13681	286	1	the	the	DET
fuelectenerg-13681	286	2	average	average	ADJ
fuelectenerg-13681	286	3	sinr	sinr	NOUN
fuelectenerg-13681	286	4	difference	difference	NOUN
fuelectenerg-13681	286	5	between	between	ADP
fuelectenerg-13681	286	6	the	the	DET
fuelectenerg-13681	286	7	best	good	ADJ
fuelectenerg-13681	286	8	traditional	traditional	ADJ
fuelectenerg-13681	286	9	algorithm	algorithm	NOUN
fuelectenerg-13681	286	10	(	(	PUNCT
fuelectenerg-13681	286	11	rls	rls	PROPN
fuelectenerg-13681	286	12	)	)	PUNCT
fuelectenerg-13681	286	13	and	and	CCONJ
fuelectenerg-13681	286	14	the	the	DET
fuelectenerg-13681	286	15	best	good	ADJ
fuelectenerg-13681	286	16	deep	deep	ADJ
fuelectenerg-13681	286	17	learning	learning	NOUN
fuelectenerg-13681	286	18	approach	approach	NOUN
fuelectenerg-13681	286	19	(	(	PUNCT
fuelectenerg-13681	286	20	rnnlstm	rnnlstm	NOUN
fuelectenerg-13681	286	21	)	)	PUNCT
fuelectenerg-13681	286	22	is	be	AUX
fuelectenerg-13681	286	23	less	less	ADJ
fuelectenerg-13681	286	24	than	than	ADP
fuelectenerg-13681	286	25	0.6	0.6	NUM
fuelectenerg-13681	286	26	db	db	PROPN
fuelectenerg-13681	286	27	(	(	PUNCT
fuelectenerg-13681	286	28	p	p	X
fuelectenerg-13681	286	29	=	=	NOUN
fuelectenerg-13681	286	30	0.08	0.08	NUM
fuelectenerg-13681	286	31	)	)	PUNCT
fuelectenerg-13681	286	32	.	.	PUNCT
fuelectenerg-13681	287	1	as	as	ADP
fuelectenerg-13681	287	2	mobility	mobility	NOUN
fuelectenerg-13681	287	3	increases	increase	NOUN
fuelectenerg-13681	287	4	,	,	PUNCT
fuelectenerg-13681	287	5	deep	deep	ADJ
fuelectenerg-13681	287	6	learning	learning	NOUN
fuelectenerg-13681	287	7	approaches	approach	NOUN
fuelectenerg-13681	287	8	(	(	PUNCT
fuelectenerg-13681	287	9	especially	especially	ADV
fuelectenerg-13681	287	10	rnn	rnn	NOUN
fuelectenerg-13681	287	11	-	-	PUNCT
fuelectenerg-13681	287	12	lstm	lstm	PROPN
fuelectenerg-13681	287	13	and	and	CCONJ
fuelectenerg-13681	287	14	mbdl	mbdl	NOUN
fuelectenerg-13681	287	15	)	)	PUNCT
fuelectenerg-13681	287	16	significantly	significantly	ADV
fuelectenerg-13681	287	17	outperform	outperform	VERB
fuelectenerg-13681	287	18	traditional	traditional	ADJ
fuelectenerg-13681	287	19	methods	method	NOUN
fuelectenerg-13681	287	20	.	.	PUNCT
fuelectenerg-13681	288	1	at	at	ADP
fuelectenerg-13681	288	2	vehicular	vehicular	ADJ
fuelectenerg-13681	288	3	speed	speed	NOUN
fuelectenerg-13681	288	4	(	(	PUNCT
fuelectenerg-13681	288	5	60	60	NUM
fuelectenerg-13681	288	6	km	km	NOUN
fuelectenerg-13681	288	7	/	/	SYM
fuelectenerg-13681	288	8	h	h	NOUN
fuelectenerg-13681	288	9	)	)	PUNCT
fuelectenerg-13681	288	10	,	,	PUNCT
fuelectenerg-13681	288	11	rnn	rnn	PROPN
fuelectenerg-13681	288	12	-	-	PUNCT
fuelectenerg-13681	288	13	lstm	lstm	PROPN
fuelectenerg-13681	288	14	achieves	achieve	VERB
fuelectenerg-13681	288	15	2.3	2.3	NUM
fuelectenerg-13681	288	16	db	db	ADP
fuelectenerg-13681	288	17	higher	high	ADJ
fuelectenerg-13681	288	18	sinr	sinr	NOUN
fuelectenerg-13681	288	19	than	than	ADP
fuelectenerg-13681	288	20	rls	rls	PROPN
fuelectenerg-13681	288	21	(	(	PUNCT
fuelectenerg-13681	288	22	p	p	X
fuelectenerg-13681	288	23	<	<	X
fuelectenerg-13681	288	24	0.01	0.01	NUM
fuelectenerg-13681	288	25	)	)	PUNCT
fuelectenerg-13681	288	26	.	.	PUNCT
fuelectenerg-13681	289	1	at	at	ADP
fuelectenerg-13681	289	2	high	high	ADJ
fuelectenerg-13681	289	3	speed	speed	NOUN
fuelectenerg-13681	289	4	(	(	PUNCT
fuelectenerg-13681	289	5	120	120	NUM
fuelectenerg-13681	289	6	km	km	NOUN
fuelectenerg-13681	289	7	/	/	SYM
fuelectenerg-13681	289	8	h	h	NOUN
fuelectenerg-13681	289	9	)	)	PUNCT
fuelectenerg-13681	289	10	,	,	PUNCT
fuelectenerg-13681	289	11	rnn	rnn	PROPN
fuelectenerg-13681	289	12	-	-	PUNCT
fuelectenerg-13681	289	13	lstm	lstm	PROPN
fuelectenerg-13681	289	14	maintains	maintain	VERB
fuelectenerg-13681	289	15	sinr	sinr	NOUN
fuelectenerg-13681	289	16	within	within	ADP
fuelectenerg-13681	289	17	1.2	1.2	NUM
fuelectenerg-13681	289	18	db	db	NOUN
fuelectenerg-13681	289	19	of	of	ADP
fuelectenerg-13681	289	20	the	the	DET
fuelectenerg-13681	289	21	optimum	optimum	NOUN
fuelectenerg-13681	289	22	,	,	PUNCT
fuelectenerg-13681	289	23	while	while	SCONJ
fuelectenerg-13681	289	24	traditional	traditional	ADJ
fuelectenerg-13681	289	25	algorithms	algorithm	NOUN
fuelectenerg-13681	289	26	degrade	degrade	VERB
fuelectenerg-13681	289	27	by	by	ADP
fuelectenerg-13681	289	28	3.5–5.2	3.5–5.2	NUM
fuelectenerg-13681	289	29	db	db	PROPN
fuelectenerg-13681	289	30	(	(	PUNCT
fuelectenerg-13681	289	31	p	p	X
fuelectenerg-13681	289	32	<	<	X
fuelectenerg-13681	289	33	0.001	0.001	NUM
fuelectenerg-13681	289	34	)	)	PUNCT
fuelectenerg-13681	289	35	.	.	PUNCT
fuelectenerg-13681	290	1	in	in	ADP
fuelectenerg-13681	290	2	extreme	extreme	ADJ
fuelectenerg-13681	290	3	mobility	mobility	NOUN
fuelectenerg-13681	290	4	scenarios	scenario	NOUN
fuelectenerg-13681	290	5	(	(	PUNCT
fuelectenerg-13681	290	6	500	500	NUM
fuelectenerg-13681	290	7	km	km	NOUN
fuelectenerg-13681	290	8	/	/	SYM
fuelectenerg-13681	290	9	h	h	NOUN
fuelectenerg-13681	290	10	)	)	PUNCT
fuelectenerg-13681	290	11	,	,	PUNCT
fuelectenerg-13681	290	12	deep	deep	ADJ
fuelectenerg-13681	290	13	learning	learning	NOUN
fuelectenerg-13681	290	14	achieves	achieve	VERB
fuelectenerg-13681	290	15	23.7	23.7	NUM
fuelectenerg-13681	290	16	%	%	NOUN
fuelectenerg-13681	290	17	higher	high	ADJ
fuelectenerg-13681	290	18	average	average	ADJ
fuelectenerg-13681	290	19	sinr	sinr	NOUN
fuelectenerg-13681	290	20	compared	compare	VERB
fuelectenerg-13681	290	21	to	to	ADP
fuelectenerg-13681	290	22	the	the	DET
fuelectenerg-13681	290	23	best	good	ADJ
fuelectenerg-13681	290	24	traditional	traditional	ADJ
fuelectenerg-13681	290	25	algorithm	algorithm	NOUN
fuelectenerg-13681	290	26	(	(	PUNCT
fuelectenerg-13681	290	27	p	p	X
fuelectenerg-13681	290	28	<	<	X
fuelectenerg-13681	290	29	0.001	0.001	NUM
fuelectenerg-13681	290	30	)	)	PUNCT
fuelectenerg-13681	290	31	.	.	PUNCT
fuelectenerg-13681	291	1	this	this	DET
fuelectenerg-13681	291	2	superior	superior	ADJ
fuelectenerg-13681	291	3	performance	performance	NOUN
fuelectenerg-13681	291	4	can	can	AUX
fuelectenerg-13681	291	5	be	be	AUX
fuelectenerg-13681	291	6	attributed	attribute	VERB
fuelectenerg-13681	291	7	to	to	ADP
fuelectenerg-13681	291	8	the	the	DET
fuelectenerg-13681	291	9	ability	ability	NOUN
fuelectenerg-13681	291	10	of	of	ADP
fuelectenerg-13681	291	11	deep	deep	ADJ
fuelectenerg-13681	291	12	learning	learning	NOUN
fuelectenerg-13681	291	13	models	model	NOUN
fuelectenerg-13681	291	14	to	to	PART
fuelectenerg-13681	291	15	learn	learn	VERB
fuelectenerg-13681	291	16	predictive	predictive	ADJ
fuelectenerg-13681	291	17	features	feature	NOUN
fuelectenerg-13681	291	18	from	from	ADP
fuelectenerg-13681	291	19	temporal	temporal	ADJ
fuelectenerg-13681	291	20	patterns	pattern	NOUN
fuelectenerg-13681	291	21	in	in	ADP
fuelectenerg-13681	291	22	channel	channel	NOUN
fuelectenerg-13681	291	23	evolution	evolution	NOUN
fuelectenerg-13681	291	24	.	.	PUNCT
fuelectenerg-13681	292	1	theoretical	theoretical	ADJ
fuelectenerg-13681	292	2	analysis	analysis	NOUN
fuelectenerg-13681	292	3	of	of	ADP
fuelectenerg-13681	292	4	tracking	tracking	NOUN
fuelectenerg-13681	292	5	performance	performance	NOUN
fuelectenerg-13681	292	6	in	in	ADP
fuelectenerg-13681	292	7	timevarying	timevarye	VERB
fuelectenerg-13681	292	8	channels	channel	NOUN
fuelectenerg-13681	292	9	shows	show	VERB
fuelectenerg-13681	292	10	that	that	SCONJ
fuelectenerg-13681	292	11	for	for	ADP
fuelectenerg-13681	292	12	a	a	DET
fuelectenerg-13681	292	13	channel	channel	NOUN
fuelectenerg-13681	292	14	with	with	ADP
fuelectenerg-13681	292	15	temporal	temporal	ADJ
fuelectenerg-13681	292	16	correlation	correlation	NOUN
fuelectenerg-13681	292	17	coefficient	coefficient	NOUN
fuelectenerg-13681	292	18			X
fuelectenerg-13681	292	19	and	and	CCONJ
fuelectenerg-13681	292	20	algorithm	algorithm	NOUN
fuelectenerg-13681	292	21	convergence	convergence	NOUN
fuelectenerg-13681	292	22	rate	rate	NOUN
fuelectenerg-13681	292	23			NOUN
fuelectenerg-13681	292	24	the	the	DET
fuelectenerg-13681	292	25	steady	steady	ADJ
fuelectenerg-13681	292	26	-	-	PUNCT
fuelectenerg-13681	292	27	state	state	NOUN
fuelectenerg-13681	292	28	sinr	sinr	NOUN
fuelectenerg-13681	292	29	loss	loss	NOUN
fuelectenerg-13681	292	30	is	be	AUX
fuelectenerg-13681	292	31	approximately	approximately	ADV
fuelectenerg-13681	292	32	proportional	proportional	ADJ
fuelectenerg-13681	292	33	to	to	ADP
fuelectenerg-13681	292	34	(	(	PUNCT
fuelectenerg-13681	292	35	1	1	NUM
fuelectenerg-13681	292	36	−	−	NOUN
fuelectenerg-13681	292	37			NOUN
fuelectenerg-13681	292	38	)	)	PUNCT
fuelectenerg-13681	292	39	/	/	SYM
fuelectenerg-13681	292	40	.	.	NOUN
fuelectenerg-13681	292	41	this	this	PRON
fuelectenerg-13681	292	42	explains	explain	VERB
fuelectenerg-13681	292	43	why	why	SCONJ
fuelectenerg-13681	292	44	algorithms	algorithm	NOUN
fuelectenerg-13681	292	45	with	with	ADP
fuelectenerg-13681	292	46	faster	fast	ADJ
fuelectenerg-13681	292	47	convergence	convergence	NOUN
fuelectenerg-13681	292	48	(	(	PUNCT
fuelectenerg-13681	292	49	deep	deep	ADJ
fuelectenerg-13681	292	50	learning	learning	NOUN
fuelectenerg-13681	292	51	during	during	ADP
fuelectenerg-13681	292	52	inference	inference	NOUN
fuelectenerg-13681	292	53	)	)	PUNCT
fuelectenerg-13681	292	54	outperform	outperform	VERB
fuelectenerg-13681	292	55	slower	slow	ADJ
fuelectenerg-13681	292	56	-	-	PUNCT
fuelectenerg-13681	292	57	converging	converge	VERB
fuelectenerg-13681	292	58	traditional	traditional	ADJ
fuelectenerg-13681	292	59	methods	method	NOUN
fuelectenerg-13681	292	60	in	in	ADP
fuelectenerg-13681	292	61	high	high	ADJ
fuelectenerg-13681	292	62	-	-	PUNCT
fuelectenerg-13681	292	63	mobility	mobility	NOUN
fuelectenerg-13681	292	64	scenarios	scenario	NOUN
fuelectenerg-13681	292	65	.	.	PUNCT
fuelectenerg-13681	293	1	the	the	DET
fuelectenerg-13681	293	2	convergence	convergence	NOUN
fuelectenerg-13681	293	3	time	time	NOUN
fuelectenerg-13681	293	4	measurements	measurement	NOUN
fuelectenerg-13681	293	5	following	follow	VERB
fuelectenerg-13681	293	6	abrupt	abrupt	ADJ
fuelectenerg-13681	293	7	channel	channel	NOUN
fuelectenerg-13681	293	8	changes	change	NOUN
fuelectenerg-13681	293	9	are	be	AUX
fuelectenerg-13681	293	10	summarized	summarize	VERB
fuelectenerg-13681	293	11	in	in	ADP
fuelectenerg-13681	293	12	table	table	NOUN
fuelectenerg-13681	293	13	3	3	NUM
fuelectenerg-13681	293	14	,	,	PUNCT
fuelectenerg-13681	293	15	clearly	clearly	ADV
fuelectenerg-13681	293	16	demonstrating	demonstrate	VERB
fuelectenerg-13681	293	17	the	the	DET
fuelectenerg-13681	293	18	speed	speed	NOUN
fuelectenerg-13681	293	19	advantage	advantage	NOUN
fuelectenerg-13681	293	20	of	of	ADP
fuelectenerg-13681	293	21	deep	deep	ADJ
fuelectenerg-13681	293	22	learning	learning	NOUN
fuelectenerg-13681	293	23	approaches	approach	NOUN
fuelectenerg-13681	293	24	.	.	PUNCT
fuelectenerg-13681	294	1	table	table	NOUN
fuelectenerg-13681	294	2	3	3	NUM
fuelectenerg-13681	294	3	convergence	convergence	NOUN
fuelectenerg-13681	294	4	time	time	NOUN
fuelectenerg-13681	294	5	following	follow	VERB
fuelectenerg-13681	294	6	abrupt	abrupt	ADJ
fuelectenerg-13681	294	7	channel	channel	NOUN
fuelectenerg-13681	294	8	changes	change	NOUN
fuelectenerg-13681	294	9	2*algorithm	2*algorithm	NUM
fuelectenerg-13681	294	10	time	time	NOUN
fuelectenerg-13681	294	11	(	(	PUNCT
fuelectenerg-13681	294	12	ms	ms	NOUN
fuelectenerg-13681	294	13	)	)	PUNCT
fuelectenerg-13681	294	14	to	to	ADP
fuelectenerg-13681	294	15	90	90	NUM
fuelectenerg-13681	294	16	%	%	NOUN
fuelectenerg-13681	294	17	optimal	optimal	ADJ
fuelectenerg-13681	294	18	sinr	sinr	NOUN
fuelectenerg-13681	294	19	to	to	ADP
fuelectenerg-13681	294	20	99	99	NUM
fuelectenerg-13681	294	21	%	%	NOUN
fuelectenerg-13681	294	22	optimal	optimal	ADJ
fuelectenerg-13681	294	23	sinr	sinr	NOUN
fuelectenerg-13681	294	24	lms	lm	NOUN
fuelectenerg-13681	295	1	32.5	32.5	NUM
fuelectenerg-13681	295	2			PROPN
fuelectenerg-13681	295	3	.1	.1	NUM
fuelectenerg-13681	296	1	78.4	78.4	NUM
fuelectenerg-13681	296	2			PROPN
fuelectenerg-13681	296	3	.6	.6	NUM
fuelectenerg-13681	296	4	rls	rls	PROPN
fuelectenerg-13681	296	5	18.7	18.7	NUM
fuelectenerg-13681	296	6			PROPN
fuelectenerg-13681	296	7	.5	.5	NUM
fuelectenerg-13681	297	1	42.6	42.6	NUM
fuelectenerg-13681	297	2			PROPN
fuelectenerg-13681	297	3	.2	.2	NUM
fuelectenerg-13681	297	4	mvdr	mvdr	VERB
fuelectenerg-13681	297	5	15.3	15.3	NUM
fuelectenerg-13681	297	6			PROPN
fuelectenerg-13681	298	1	.2	.2	NUM
fuelectenerg-13681	298	2	38.5	38.5	NUM
fuelectenerg-13681	299	1			PROPN
fuelectenerg-13681	299	2	.9	.9	NUM
fuelectenerg-13681	299	3	socp	socp	NOUN
fuelectenerg-13681	299	4	24.6	24.6	NUM
fuelectenerg-13681	300	1			PROPN
fuelectenerg-13681	301	1	.9	.9	NUM
fuelectenerg-13681	302	1	52.8	52.8	NUM
fuelectenerg-13681	302	2			PROPN
fuelectenerg-13681	302	3	.7	.7	PROPN
fuelectenerg-13681	302	4	fc	fc	PROPN
fuelectenerg-13681	302	5	-	-	PUNCT
fuelectenerg-13681	302	6	dnn	dnn	PROPN
fuelectenerg-13681	302	7	3.8	3.8	NUM
fuelectenerg-13681	302	8			PROPN
fuelectenerg-13681	302	9	.3	.3	NUM
fuelectenerg-13681	302	10	8.6	8.6	NUM
fuelectenerg-13681	303	1			PROPN
fuelectenerg-13681	303	2	.7	.7	PROPN
fuelectenerg-13681	303	3	cnn	cnn	PROPN
fuelectenerg-13681	303	4	3.2	3.2	NUM
fuelectenerg-13681	304	1			PROPN
fuelectenerg-13681	304	2	.3	.3	NUM
fuelectenerg-13681	304	3	7.5	7.5	NUM
fuelectenerg-13681	305	1			PROPN
fuelectenerg-13681	305	2	.6	.6	NUM
fuelectenerg-13681	305	3	rnn	rnn	PROPN
fuelectenerg-13681	305	4	-	-	PUNCT
fuelectenerg-13681	305	5	lstm	lstm	ADJ
fuelectenerg-13681	305	6	2.4	2.4	NUM
fuelectenerg-13681	305	7			PROPN
fuelectenerg-13681	305	8	.2	.2	NUM
fuelectenerg-13681	305	9	5.8	5.8	NUM
fuelectenerg-13681	305	10			PROPN
fuelectenerg-13681	305	11	.5	.5	NUM
fuelectenerg-13681	305	12	mbdl	mbdl	NOUN
fuelectenerg-13681	305	13	4.1	4.1	NUM
fuelectenerg-13681	306	1			PROPN
fuelectenerg-13681	306	2	.4	.4	NUM
fuelectenerg-13681	306	3	9.2	9.2	NUM
fuelectenerg-13681	307	1			PROPN
fuelectenerg-13681	307	2	.8	.8	NUM
fuelectenerg-13681	308	1	deep	deep	ADJ
fuelectenerg-13681	308	2	learning	learning	NOUN
fuelectenerg-13681	308	3	approaches	approach	NOUN
fuelectenerg-13681	308	4	demonstrate	demonstrate	VERB
fuelectenerg-13681	308	5	significantly	significantly	ADV
fuelectenerg-13681	308	6	faster	fast	ADJ
fuelectenerg-13681	308	7	convergence	convergence	NOUN
fuelectenerg-13681	308	8	following	follow	VERB
fuelectenerg-13681	308	9	abrupt	abrupt	ADJ
fuelectenerg-13681	308	10	channel	channel	NOUN
fuelectenerg-13681	308	11	changes	change	NOUN
fuelectenerg-13681	308	12	(	(	PUNCT
fuelectenerg-13681	308	13	p	p	X
fuelectenerg-13681	308	14	<	<	X
fuelectenerg-13681	308	15	0.001	0.001	NUM
fuelectenerg-13681	308	16	)	)	PUNCT
fuelectenerg-13681	308	17	.	.	PUNCT
fuelectenerg-13681	309	1	the	the	DET
fuelectenerg-13681	309	2	fastest	fast	ADJ
fuelectenerg-13681	309	3	traditional	traditional	ADJ
fuelectenerg-13681	309	4	algorithm	algorithm	NOUN
fuelectenerg-13681	309	5	(	(	PUNCT
fuelectenerg-13681	309	6	mvdr	mvdr	NOUN
fuelectenerg-13681	309	7	)	)	PUNCT
fuelectenerg-13681	309	8	requires	require	VERB
fuelectenerg-13681	309	9	15.3	15.3	NUM
fuelectenerg-13681	309	10	ms	ms	NOUN
fuelectenerg-13681	309	11	to	to	PART
fuelectenerg-13681	309	12	reach	reach	VERB
fuelectenerg-13681	309	13	90	90	NUM
fuelectenerg-13681	309	14	%	%	NOUN
fuelectenerg-13681	309	15	of	of	ADP
fuelectenerg-13681	309	16	optimal	optimal	ADJ
fuelectenerg-13681	309	17	sinr	sinr	NOUN
fuelectenerg-13681	309	18	,	,	PUNCT
fuelectenerg-13681	309	19	while	while	SCONJ
fuelectenerg-13681	309	20	rnn	rnn	NOUN
fuelectenerg-13681	309	21	-	-	PUNCT
fuelectenerg-13681	309	22	lstm	lstm	PROPN
fuelectenerg-13681	309	23	achieves	achieve	VERB
fuelectenerg-13681	309	24	the	the	DET
fuelectenerg-13681	309	25	same	same	ADJ
fuelectenerg-13681	309	26	level	level	NOUN
fuelectenerg-13681	309	27	in	in	ADP
fuelectenerg-13681	309	28	just	just	ADV
fuelectenerg-13681	309	29	2.4	2.4	NUM
fuelectenerg-13681	309	30	ms	ms	NOUN
fuelectenerg-13681	309	31	–	–	PUNCT
fuelectenerg-13681	309	32	a	a	DET
fuelectenerg-13681	309	33	6.4×	6.4×	NUM
fuelectenerg-13681	309	34	improvement	improvement	NOUN
fuelectenerg-13681	309	35	.	.	PUNCT
fuelectenerg-13681	310	1	to	to	PART
fuelectenerg-13681	310	2	assess	assess	VERB
fuelectenerg-13681	310	3	robustness	robustness	NOUN
fuelectenerg-13681	310	4	to	to	AUX
fuelectenerg-13681	310	5	channel	channel	VERB
fuelectenerg-13681	310	6	estimation	estimation	NOUN
fuelectenerg-13681	310	7	errors	error	NOUN
fuelectenerg-13681	310	8	,	,	PUNCT
fuelectenerg-13681	310	9	we	we	PRON
fuelectenerg-13681	310	10	introduced	introduce	VERB
fuelectenerg-13681	310	11	controlled	control	VERB
fuelectenerg-13681	310	12	perturbations	perturbation	NOUN
fuelectenerg-13681	310	13	to	to	ADP
fuelectenerg-13681	310	14	the	the	DET
fuelectenerg-13681	310	15	channel	channel	NOUN
fuelectenerg-13681	310	16	matrices	matrix	NOUN
fuelectenerg-13681	310	17	.	.	PUNCT
fuelectenerg-13681	311	1	with	with	ADP
fuelectenerg-13681	311	2	10	10	NUM
fuelectenerg-13681	311	3	%	%	NOUN
fuelectenerg-13681	311	4	channel	channel	NOUN
fuelectenerg-13681	311	5	estimation	estimation	NOUN
fuelectenerg-13681	311	6	error	error	NOUN
fuelectenerg-13681	311	7	,	,	PUNCT
fuelectenerg-13681	311	8	cnn	cnn	PROPN
fuelectenerg-13681	311	9	and	and	CCONJ
fuelectenerg-13681	311	10	rnn	rnn	PROPN
fuelectenerg-13681	311	11	-	-	PUNCT
fuelectenerg-13681	311	12	lstm	lstm	NOUN
fuelectenerg-13681	311	13	lose	lose	VERB
fuelectenerg-13681	311	14	only	only	ADV
fuelectenerg-13681	311	15	1.8	1.8	NUM
fuelectenerg-13681	311	16	db	db	NOUN
fuelectenerg-13681	311	17	and	and	CCONJ
fuelectenerg-13681	311	18	2.1	2.1	NUM
fuelectenerg-13681	311	19	db	db	NOUN
fuelectenerg-13681	311	20	sinr	sinr	NOUN
fuelectenerg-13681	311	21	respectively	respectively	ADV
fuelectenerg-13681	311	22	,	,	PUNCT
fuelectenerg-13681	311	23	compared	compare	VERB
fuelectenerg-13681	311	24	to	to	ADP
fuelectenerg-13681	311	25	3.5	3.5	NUM
fuelectenerg-13681	311	26	db	db	NOUN
fuelectenerg-13681	311	27	for	for	ADP
fuelectenerg-13681	311	28	mvdr	mvdr	NOUN
fuelectenerg-13681	311	29	and	and	CCONJ
fuelectenerg-13681	311	30	3.2	3.2	NUM
fuelectenerg-13681	311	31	db	db	NOUN
fuelectenerg-13681	311	32	for	for	ADP
fuelectenerg-13681	311	33	rls	rls	PROPN
fuelectenerg-13681	311	34	(	(	PUNCT
fuelectenerg-13681	311	35	p	p	X
fuelectenerg-13681	311	36	<	<	X
fuelectenerg-13681	311	37	0.01	0.01	NUM
fuelectenerg-13681	311	38	)	)	PUNCT
fuelectenerg-13681	311	39	.	.	PUNCT
fuelectenerg-13681	312	1	this	this	PRON
fuelectenerg-13681	312	2	demonstrates	demonstrate	VERB
fuelectenerg-13681	312	3	the	the	DET
fuelectenerg-13681	312	4	superior	superior	ADJ
fuelectenerg-13681	312	5	robustness	robustness	NOUN
fuelectenerg-13681	312	6	of	of	ADP
fuelectenerg-13681	312	7	deep	deep	ADJ
fuelectenerg-13681	312	8	learning	learning	NOUN
fuelectenerg-13681	312	9	approaches	approach	NOUN
fuelectenerg-13681	312	10	to	to	ADP
fuelectenerg-13681	312	11	imperfect	imperfect	ADJ
fuelectenerg-13681	312	12	channel	channel	NOUN
fuelectenerg-13681	312	13	state	state	NOUN
fuelectenerg-13681	312	14	information	information	NOUN
fuelectenerg-13681	312	15	.	.	PUNCT
fuelectenerg-13681	313	1	the	the	DET
fuelectenerg-13681	313	2	hardware	hardware	NOUN
fuelectenerg-13681	313	3	validation	validation	NOUN
fuelectenerg-13681	313	4	on	on	ADP
fuelectenerg-13681	313	5	usrp	usrp	NOUN
fuelectenerg-13681	313	6	x310	x310	PROPN
fuelectenerg-13681	313	7	platforms	platform	NOUN
fuelectenerg-13681	313	8	confirms	confirm	VERB
fuelectenerg-13681	313	9	these	these	DET
fuelectenerg-13681	313	10	simulation	simulation	NOUN
fuelectenerg-13681	313	11	results	result	VERB
fuelectenerg-13681	313	12	.	.	PUNCT
fuelectenerg-13681	314	1	in	in	ADP
fuelectenerg-13681	314	2	a	a	DET
fuelectenerg-13681	314	3	controlled	control	VERB
fuelectenerg-13681	314	4	laboratory	laboratory	NOUN
fuelectenerg-13681	314	5	environment	environment	NOUN
fuelectenerg-13681	314	6	with	with	ADP
fuelectenerg-13681	314	7	programmable	programmable	ADJ
fuelectenerg-13681	314	8	channel	channel	NOUN
fuelectenerg-13681	314	9	emulation	emulation	NOUN
fuelectenerg-13681	314	10	,	,	PUNCT
fuelectenerg-13681	314	11	we	we	PRON
fuelectenerg-13681	314	12	measured	measure	VERB
fuelectenerg-13681	314	13	convergence	convergence	NOUN
fuelectenerg-13681	314	14	times	time	NOUN
fuelectenerg-13681	314	15	within	within	ADP
fuelectenerg-13681	314	16	12	12	NUM
fuelectenerg-13681	314	17	%	%	NOUN
fuelectenerg-13681	314	18	of	of	ADP
fuelectenerg-13681	314	19	the	the	DET
fuelectenerg-13681	314	20	simulation	simulation	NOUN
fuelectenerg-13681	314	21	predictions	prediction	NOUN
fuelectenerg-13681	314	22	.	.	PUNCT
fuelectenerg-13681	315	1	the	the	DET
fuelectenerg-13681	315	2	relative	relative	ADJ
fuelectenerg-13681	315	3	performance	performance	NOUN
fuelectenerg-13681	315	4	ranking	ranking	NOUN
fuelectenerg-13681	315	5	of	of	ADP
fuelectenerg-13681	315	6	algorithms	algorithm	NOUN
fuelectenerg-13681	315	7	remained	remain	VERB
fuelectenerg-13681	315	8	consistent	consistent	ADJ
fuelectenerg-13681	315	9	across	across	ADP
fuelectenerg-13681	315	10	simulation	simulation	NOUN
fuelectenerg-13681	315	11	and	and	CCONJ
fuelectenerg-13681	315	12	hardware	hardware	NOUN
fuelectenerg-13681	315	13	implementation	implementation	NOUN
fuelectenerg-13681	315	14	.	.	PUNCT
fuelectenerg-13681	316	1	212	212	NUM
fuelectenerg-13681	316	2	m.	m.	NOUN
fuelectenerg-13681	316	3	mohamed	mohamed	PROPN
fuelectenerg-13681	316	4	,	,	PUNCT
fuelectenerg-13681	316	5	f.	f.	PROPN
fuelectenerg-13681	316	6	aljuaid	aljuaid	PROPN
fuelectenerg-13681	316	7	,	,	PUNCT
fuelectenerg-13681	316	8	mhd	mhd	PROPN
fuelectenerg-13681	316	9	w.	w.	PROPN
fuelectenerg-13681	316	10	koubeisi	koubeisi	VERB
fuelectenerg-13681	316	11	4.3	4.3	NUM
fuelectenerg-13681	316	12	.	.	PUNCT
fuelectenerg-13681	317	1	computational	computational	ADJ
fuelectenerg-13681	317	2	complexity	complexity	NOUN
fuelectenerg-13681	317	3	and	and	CCONJ
fuelectenerg-13681	317	4	resource	resource	NOUN
fuelectenerg-13681	317	5	requirements	requirement	NOUN
fuelectenerg-13681	317	6	fig	fig	NOUN
fuelectenerg-13681	317	7	.	.	PUNCT
fuelectenerg-13681	318	1	3	3	NUM
fuelectenerg-13681	318	2	computational	computational	ADJ
fuelectenerg-13681	318	3	complexity	complexity	NOUN
fuelectenerg-13681	318	4	comparison	comparison	NOUN
fuelectenerg-13681	318	5	between	between	ADP
fuelectenerg-13681	318	6	traditional	traditional	ADJ
fuelectenerg-13681	318	7	and	and	CCONJ
fuelectenerg-13681	318	8	deep	deep	ADJ
fuelectenerg-13681	318	9	learning	learning	NOUN
fuelectenerg-13681	318	10	approaches	approach	NOUN
fuelectenerg-13681	318	11	across	across	ADP
fuelectenerg-13681	318	12	different	different	ADJ
fuelectenerg-13681	318	13	mimo	mimo	PROPN
fuelectenerg-13681	318	14	configurations	configuration	NOUN
fuelectenerg-13681	318	15	for	for	ADP
fuelectenerg-13681	318	16	small	small	ADJ
fuelectenerg-13681	318	17	mimo	mimo	PROPN
fuelectenerg-13681	318	18	systems	system	NOUN
fuelectenerg-13681	318	19	(	(	PUNCT
fuelectenerg-13681	318	20	4×4	4×4	NUM
fuelectenerg-13681	318	21	)	)	PUNCT
fuelectenerg-13681	318	22	,	,	PUNCT
fuelectenerg-13681	318	23	lms	lms	PROPN
fuelectenerg-13681	318	24	is	be	AUX
fuelectenerg-13681	318	25	the	the	DET
fuelectenerg-13681	318	26	most	most	ADV
fuelectenerg-13681	318	27	computationally	computationally	ADV
fuelectenerg-13681	318	28	efficient	efficient	ADJ
fuelectenerg-13681	318	29	algorithm	algorithm	NOUN
fuelectenerg-13681	318	30	with	with	ADP
fuelectenerg-13681	318	31	approximately	approximately	ADV
fuelectenerg-13681	318	32	2,500	2,500	NUM
fuelectenerg-13681	318	33	flops	flop	NOUN
fuelectenerg-13681	318	34	per	per	ADP
fuelectenerg-13681	318	35	beamforming	beamforme	VERB
fuelectenerg-13681	318	36	update	update	NOUN
fuelectenerg-13681	318	37	.	.	PUNCT
fuelectenerg-13681	319	1	as	as	ADP
fuelectenerg-13681	319	2	the	the	DET
fuelectenerg-13681	319	3	mimo	mimo	PROPN
fuelectenerg-13681	319	4	size	size	NOUN
fuelectenerg-13681	319	5	increases	increase	NOUN
fuelectenerg-13681	319	6	,	,	PUNCT
fuelectenerg-13681	319	7	the	the	DET
fuelectenerg-13681	319	8	complexity	complexity	NOUN
fuelectenerg-13681	319	9	of	of	ADP
fuelectenerg-13681	319	10	traditional	traditional	ADJ
fuelectenerg-13681	319	11	methods	method	NOUN
fuelectenerg-13681	319	12	grows	grow	VERB
fuelectenerg-13681	319	13	rapidly	rapidly	ADV
fuelectenerg-13681	319	14	,	,	PUNCT
fuelectenerg-13681	319	15	with	with	ADP
fuelectenerg-13681	319	16	socp	socp	NOUN
fuelectenerg-13681	319	17	scaling	scale	VERB
fuelectenerg-13681	319	18	as	as	ADP
fuelectenerg-13681	319	19	o(n³	o(n³	PROPN
fuelectenerg-13681	319	20	)	)	PUNCT
fuelectenerg-13681	319	21	,	,	PUNCT
fuelectenerg-13681	319	22	where	where	SCONJ
fuelectenerg-13681	319	23	n	n	PRON
fuelectenerg-13681	319	24	is	be	AUX
fuelectenerg-13681	319	25	the	the	DET
fuelectenerg-13681	319	26	number	number	NOUN
fuelectenerg-13681	319	27	of	of	ADP
fuelectenerg-13681	319	28	antennas	antenna	NOUN
fuelectenerg-13681	319	29	.	.	PUNCT
fuelectenerg-13681	320	1	in	in	ADP
fuelectenerg-13681	320	2	contrast	contrast	NOUN
fuelectenerg-13681	320	3	,	,	PUNCT
fuelectenerg-13681	320	4	deep	deep	ADJ
fuelectenerg-13681	320	5	learning	learning	NOUN
fuelectenerg-13681	320	6	approaches	approach	NOUN
fuelectenerg-13681	320	7	maintain	maintain	VERB
fuelectenerg-13681	320	8	relatively	relatively	ADV
fuelectenerg-13681	320	9	stable	stable	ADJ
fuelectenerg-13681	320	10	inference	inference	NOUN
fuelectenerg-13681	320	11	complexity	complexity	NOUN
fuelectenerg-13681	320	12	once	once	ADV
fuelectenerg-13681	320	13	trained	train	VERB
fuelectenerg-13681	320	14	.	.	PUNCT
fuelectenerg-13681	321	1	for	for	ADP
fuelectenerg-13681	321	2	massive	massive	ADJ
fuelectenerg-13681	321	3	mimo	mimo	PROPN
fuelectenerg-13681	321	4	configurations	configuration	NOUN
fuelectenerg-13681	321	5	(	(	PUNCT
fuelectenerg-13681	321	6	64×64	64×64	NUM
fuelectenerg-13681	321	7	,	,	PUNCT
fuelectenerg-13681	321	8	128×128	128×128	NUM
fuelectenerg-13681	321	9	)	)	PUNCT
fuelectenerg-13681	321	10	,	,	PUNCT
fuelectenerg-13681	321	11	deep	deep	ADJ
fuelectenerg-13681	321	12	learning	learning	NOUN
fuelectenerg-13681	321	13	requires	require	VERB
fuelectenerg-13681	321	14	65–85	65–85	NUM
fuelectenerg-13681	321	15	%	%	NOUN
fuelectenerg-13681	321	16	fewer	few	ADJ
fuelectenerg-13681	321	17	flops	flop	NOUN
fuelectenerg-13681	321	18	compared	compare	VERB
fuelectenerg-13681	321	19	to	to	ADP
fuelectenerg-13681	321	20	traditional	traditional	ADJ
fuelectenerg-13681	321	21	optimization	optimization	NOUN
fuelectenerg-13681	321	22	algorithms	algorithm	NOUN
fuelectenerg-13681	321	23	(	(	PUNCT
fuelectenerg-13681	321	24	p	p	X
fuelectenerg-13681	321	25	<	<	X
fuelectenerg-13681	321	26	0.001	0.001	NUM
fuelectenerg-13681	321	27	)	)	PUNCT
fuelectenerg-13681	321	28	.	.	PUNCT
fuelectenerg-13681	322	1	the	the	DET
fuelectenerg-13681	322	2	asymptotic	asymptotic	ADJ
fuelectenerg-13681	322	3	complexity	complexity	NOUN
fuelectenerg-13681	322	4	analysis	analysis	NOUN
fuelectenerg-13681	322	5	is	be	AUX
fuelectenerg-13681	322	6	confirmed	confirm	VERB
fuelectenerg-13681	322	7	by	by	ADP
fuelectenerg-13681	322	8	our	our	PRON
fuelectenerg-13681	322	9	empirical	empirical	ADJ
fuelectenerg-13681	322	10	measurements	measurement	NOUN
fuelectenerg-13681	322	11	.	.	PUNCT
fuelectenerg-13681	323	1	for	for	ADP
fuelectenerg-13681	323	2	a	a	DET
fuelectenerg-13681	323	3	128×128	128×128	NUM
fuelectenerg-13681	323	4	mimo	mimo	PROPN
fuelectenerg-13681	323	5	system	system	NOUN
fuelectenerg-13681	323	6	,	,	PUNCT
fuelectenerg-13681	323	7	socp	socp	NOUN
fuelectenerg-13681	323	8	requires	require	VERB
fuelectenerg-13681	323	9	approximately	approximately	ADV
fuelectenerg-13681	323	10	6.8×10⁸	6.8×10⁸	NUM
fuelectenerg-13681	323	11	flops	flop	NOUN
fuelectenerg-13681	323	12	,	,	PUNCT
fuelectenerg-13681	323	13	compared	compare	VERB
fuelectenerg-13681	323	14	to	to	ADP
fuelectenerg-13681	323	15	9.7×10⁷	9.7×10⁷	NUM
fuelectenerg-13681	323	16	flops	flop	NOUN
fuelectenerg-13681	323	17	for	for	ADP
fuelectenerg-13681	323	18	cnn	cnn	PROPN
fuelectenerg-13681	323	19	inference	inference	PROPN
fuelectenerg-13681	323	20	–	–	PUNCT
fuelectenerg-13681	323	21	a	a	DET
fuelectenerg-13681	323	22	reduction	reduction	NOUN
fuelectenerg-13681	323	23	of	of	ADP
fuelectenerg-13681	323	24	85.7	85.7	NUM
fuelectenerg-13681	323	25	%	%	NOUN
fuelectenerg-13681	323	26	.	.	PUNCT
fuelectenerg-13681	324	1	theoretically	theoretically	ADV
fuelectenerg-13681	324	2	,	,	PUNCT
fuelectenerg-13681	324	3	the	the	DET
fuelectenerg-13681	324	4	computational	computational	ADJ
fuelectenerg-13681	324	5	complexity	complexity	NOUN
fuelectenerg-13681	324	6	of	of	ADP
fuelectenerg-13681	324	7	traditional	traditional	ADJ
fuelectenerg-13681	324	8	algorithms	algorithm	NOUN
fuelectenerg-13681	324	9	can	can	AUX
fuelectenerg-13681	324	10	be	be	AUX
fuelectenerg-13681	324	11	expressed	express	VERB
fuelectenerg-13681	324	12	as	as	ADP
fuelectenerg-13681	324	13	:	:	PUNCT
fuelectenerg-13681	324	14	lms	lm	NOUN
fuelectenerg-13681	324	15	:	:	PUNCT
fuelectenerg-13681	324	16	(	(	PUNCT
fuelectenerg-13681	324	17	)	)	PUNCT
fuelectenerg-13681	324	18	t	t	PROPN
fuelectenerg-13681	324	19	ro	ro	NOUN
fuelectenerg-13681	324	20	n	n	CCONJ
fuelectenerg-13681	324	21	n	n	PROPN
fuelectenerg-13681	324	22	(	(	PUNCT
fuelectenerg-13681	324	23	41	41	NUM
fuelectenerg-13681	324	24	)	)	PUNCT
fuelectenerg-13681	324	25	2rls	2rl	NOUN
fuelectenerg-13681	324	26	:	:	PUNCT
fuelectenerg-13681	324	27	(	(	PUNCT
fuelectenerg-13681	324	28	)	)	PUNCT
fuelectenerg-13681	324	29	t	t	PROPN
fuelectenerg-13681	324	30	ro	ro	NOUN
fuelectenerg-13681	324	31	n	n	CCONJ
fuelectenerg-13681	324	32	n	n	PROPN
fuelectenerg-13681	324	33	(	(	PUNCT
fuelectenerg-13681	324	34	42	42	NUM
fuelectenerg-13681	324	35	)	)	PUNCT
fuelectenerg-13681	324	36	2	2	NUM
fuelectenerg-13681	324	37	3mvdr	3mvdr	NUM
fuelectenerg-13681	324	38	:	:	PUNCT
fuelectenerg-13681	324	39	(	(	PUNCT
fuelectenerg-13681	324	40	)	)	PUNCT
fuelectenerg-13681	324	41	t	t	NOUN
fuelectenerg-13681	324	42	r	r	NOUN
fuelectenerg-13681	324	43	to	to	ADP
fuelectenerg-13681	324	44	n	n	PROPN
fuelectenerg-13681	324	45	n	n	PROPN
fuelectenerg-13681	324	46	n+	n+	X
fuelectenerg-13681	324	47	(	(	PUNCT
fuelectenerg-13681	324	48	43	43	NUM
fuelectenerg-13681	324	49	)	)	PUNCT
fuelectenerg-13681	324	50	3	3	NUM
fuelectenerg-13681	324	51	3socp	3socp	NUM
fuelectenerg-13681	324	52	:	:	PUNCT
fuelectenerg-13681	324	53	(	(	PUNCT
fuelectenerg-13681	324	54	)	)	PUNCT
fuelectenerg-13681	324	55	t	t	PROPN
fuelectenerg-13681	324	56	ro	ro	NOUN
fuelectenerg-13681	324	57	n	n	CCONJ
fuelectenerg-13681	324	58	n	n	PROPN
fuelectenerg-13681	324	59	(	(	PUNCT
fuelectenerg-13681	324	60	44	44	NUM
fuelectenerg-13681	324	61	)	)	PUNCT
fuelectenerg-13681	324	62	for	for	ADP
fuelectenerg-13681	324	63	deep	deep	ADJ
fuelectenerg-13681	324	64	learning	learning	NOUN
fuelectenerg-13681	324	65	approaches	approach	NOUN
fuelectenerg-13681	324	66	,	,	PUNCT
fuelectenerg-13681	324	67	the	the	DET
fuelectenerg-13681	324	68	inference	inference	NOUN
fuelectenerg-13681	324	69	complexity	complexity	NOUN
fuelectenerg-13681	324	70	is	be	AUX
fuelectenerg-13681	324	71	:	:	PUNCT
fuelectenerg-13681	324	72	2fc	2fc	PROPN
fuelectenerg-13681	324	73	dnn	dnn	PROPN
fuelectenerg-13681	324	74	:	:	PUNCT
fuelectenerg-13681	324	75	(	(	PUNCT
fuelectenerg-13681	324	76	)	)	PUNCT
fuelectenerg-13681	324	77	t	t	PROPN
fuelectenerg-13681	324	78	ro	ro	PROPN
fuelectenerg-13681	324	79	n	n	ADP
fuelectenerg-13681	324	80	n	n	PROPN
fuelectenerg-13681	324	81	h	h	NOUN
fuelectenerg-13681	324	82	h−	h−	VERB
fuelectenerg-13681	325	1	+	+	CCONJ
fuelectenerg-13681	325	2	(	(	PUNCT
fuelectenerg-13681	325	3	45	45	NUM
fuelectenerg-13681	325	4	)	)	PUNCT
fuelectenerg-13681	325	5	2cnn	2cnn	NOUN
fuelectenerg-13681	325	6	:	:	PUNCT
fuelectenerg-13681	325	7	(	(	PUNCT
fuelectenerg-13681	325	8	)	)	PUNCT
fuelectenerg-13681	325	9	t	t	PROPN
fuelectenerg-13681	325	10	ro	ro	PROPN
fuelectenerg-13681	325	11	n	n	CCONJ
fuelectenerg-13681	325	12	n	n	PROPN
fuelectenerg-13681	325	13	k	k	PROPN
fuelectenerg-13681	325	14	f	f	PROPN
fuelectenerg-13681	325	15	(	(	PUNCT
fuelectenerg-13681	325	16	46	46	NUM
fuelectenerg-13681	325	17	)	)	PUNCT
fuelectenerg-13681	326	1	2rnn	2rnn	NOUN
fuelectenerg-13681	326	2	lstm	lstm	NOUN
fuelectenerg-13681	326	3	:	:	PUNCT
fuelectenerg-13681	326	4	(	(	PUNCT
fuelectenerg-13681	326	5	)	)	PUNCT
fuelectenerg-13681	326	6	t	t	PROPN
fuelectenerg-13681	326	7	ro	ro	PROPN
fuelectenerg-13681	326	8	n	n	ADP
fuelectenerg-13681	326	9	n	n	PROPN
fuelectenerg-13681	326	10	h	h	NOUN
fuelectenerg-13681	326	11	h−	h−	VERB
fuelectenerg-13681	327	1	+	+	CCONJ
fuelectenerg-13681	327	2	(	(	PUNCT
fuelectenerg-13681	327	3	47	47	NUM
fuelectenerg-13681	327	4	)	)	PUNCT
fuelectenerg-13681	327	5	(	(	PUNCT
fuelectenerg-13681	327	6	)	)	PUNCT
fuelectenerg-13681	327	7	2mbdl	2mbdl	NUM
fuelectenerg-13681	327	8	:	:	PUNCT
fuelectenerg-13681	327	9	(	(	PUNCT
fuelectenerg-13681	327	10	)	)	PUNCT
fuelectenerg-13681	327	11	t	t	NOUN
fuelectenerg-13681	327	12	r	r	NOUN
fuelectenerg-13681	327	13	to	to	ADP
fuelectenerg-13681	327	14	i	i	PRON
fuelectenerg-13681	327	15	n	n	PROPN
fuelectenerg-13681	327	16	n	n	PROPN
fuelectenerg-13681	327	17	n+	n+	X
fuelectenerg-13681	327	18	(	(	PUNCT
fuelectenerg-13681	327	19	48	48	NUM
fuelectenerg-13681	327	20	)	)	PUNCT
fuelectenerg-13681	327	21	where	where	SCONJ
fuelectenerg-13681	327	22	h	h	NOUN
fuelectenerg-13681	327	23	is	be	AUX
fuelectenerg-13681	327	24	the	the	DET
fuelectenerg-13681	327	25	largest	large	ADJ
fuelectenerg-13681	327	26	hidden	hidden	ADJ
fuelectenerg-13681	327	27	layer	layer	NOUN
fuelectenerg-13681	327	28	size	size	NOUN
fuelectenerg-13681	327	29	,	,	PUNCT
fuelectenerg-13681	327	30	k	k	PROPN
fuelectenerg-13681	327	31	is	be	AUX
fuelectenerg-13681	327	32	the	the	DET
fuelectenerg-13681	327	33	kernel	kernel	NOUN
fuelectenerg-13681	327	34	size	size	NOUN
fuelectenerg-13681	327	35	,	,	PUNCT
fuelectenerg-13681	327	36	f	f	PROPN
fuelectenerg-13681	327	37	is	be	AUX
fuelectenerg-13681	327	38	the	the	DET
fuelectenerg-13681	327	39	num	num	ADJ
fuelectenerg-13681	327	40	ber	ber	NOUN
fuelectenerg-13681	327	41	of	of	ADP
fuelectenerg-13681	327	42	filters	filter	NOUN
fuelectenerg-13681	327	43	,	,	PUNCT
fuelectenerg-13681	327	44	and	and	CCONJ
fuelectenerg-13681	327	45	i	i	PRON
fuelectenerg-13681	327	46	is	be	AUX
fuelectenerg-13681	327	47	the	the	DET
fuelectenerg-13681	327	48	number	number	NOUN
fuelectenerg-13681	327	49	of	of	ADP
fuelectenerg-13681	327	50	iterations	iteration	NOUN
fuelectenerg-13681	327	51	/	/	SYM
fuelectenerg-13681	327	52	layers	layer	NOUN
fuelectenerg-13681	327	53	.	.	PUNCT
fuelectenerg-13681	328	1	as	as	SCONJ
fuelectenerg-13681	328	2	shown	show	VERB
fuelectenerg-13681	328	3	in	in	ADP
fuelectenerg-13681	328	4	table	table	NOUN
fuelectenerg-13681	328	5	4	4	NUM
fuelectenerg-13681	328	6	,	,	PUNCT
fuelectenerg-13681	328	7	the	the	DET
fuelectenerg-13681	328	8	execution	execution	NOUN
fuelectenerg-13681	328	9	time	time	NOUN
fuelectenerg-13681	328	10	measurements	measurement	NOUN
fuelectenerg-13681	328	11	deep	deep	ADJ
fuelectenerg-13681	328	12	learning	learning	NOUN
fuelectenerg-13681	328	13	and	and	CCONJ
fuelectenerg-13681	328	14	traditional	traditional	ADJ
fuelectenerg-13681	328	15	algorithms	algorithm	NOUN
fuelectenerg-13681	328	16	for	for	ADP
fuelectenerg-13681	328	17	mimo	mimo	PROPN
fuelectenerg-13681	328	18	beamforming	beamforme	VERB
fuelectenerg-13681	328	19	213	213	NUM
fuelectenerg-13681	328	20	across	across	ADP
fuelectenerg-13681	328	21	different	different	ADJ
fuelectenerg-13681	328	22	hardware	hardware	NOUN
fuelectenerg-13681	328	23	platforms	platform	NOUN
fuelectenerg-13681	328	24	and	and	CCONJ
fuelectenerg-13681	328	25	mimo	mimo	PROPN
fuelectenerg-13681	328	26	configurations	configuration	NOUN
fuelectenerg-13681	328	27	confirm	confirm	VERB
fuelectenerg-13681	328	28	the	the	DET
fuelectenerg-13681	328	29	computational	computational	ADJ
fuelectenerg-13681	328	30	efficiency	efficiency	NOUN
fuelectenerg-13681	328	31	of	of	ADP
fuelectenerg-13681	328	32	deep	deep	ADJ
fuelectenerg-13681	328	33	learning	learning	NOUN
fuelectenerg-13681	328	34	approaches	approach	NOUN
fuelectenerg-13681	328	35	,	,	PUNCT
fuelectenerg-13681	328	36	particularly	particularly	ADV
fuelectenerg-13681	328	37	for	for	ADP
fuelectenerg-13681	328	38	larger	large	ADJ
fuelectenerg-13681	328	39	antenna	antenna	NOUN
fuelectenerg-13681	328	40	arrays	array	VERB
fuelectenerg-13681	328	41	.	.	PUNCT
fuelectenerg-13681	329	1	table	table	NOUN
fuelectenerg-13681	329	2	4	4	NUM
fuelectenerg-13681	329	3	execution	execution	NOUN
fuelectenerg-13681	329	4	time	time	NOUN
fuelectenerg-13681	329	5	comparison	comparison	NOUN
fuelectenerg-13681	329	6	on	on	ADP
fuelectenerg-13681	329	7	different	different	ADJ
fuelectenerg-13681	329	8	hardware	hardware	NOUN
fuelectenerg-13681	329	9	platforms	platform	NOUN
fuelectenerg-13681	329	10	3*algorithm	3*algorithm	PROPN
fuelectenerg-13681	329	11	execution	execution	NOUN
fuelectenerg-13681	329	12	time	time	NOUN
fuelectenerg-13681	329	13	(	(	PUNCT
fuelectenerg-13681	329	14	ms	ms	NOUN
fuelectenerg-13681	329	15	)	)	PUNCT
fuelectenerg-13681	329	16	workstation	workstation	PROPN
fuelectenerg-13681	329	17	jetson	jetson	PROPN
fuelectenerg-13681	329	18	agx	agx	PROPN
fuelectenerg-13681	329	19	xavier	xavier	PROPN
fuelectenerg-13681	329	20	8×8	8×8	PROPN
fuelectenerg-13681	330	1	32×32	32×32	NUM
fuelectenerg-13681	330	2	8×8	8×8	NUM
fuelectenerg-13681	330	3	32×32	32×32	NUM
fuelectenerg-13681	330	4	lms	lm	NOUN
fuelectenerg-13681	330	5	0.42	0.42	NUM
fuelectenerg-13681	330	6			PROPN
fuelectenerg-13681	330	7	.03	.03	NUM
fuelectenerg-13681	330	8	3.84	3.84	NUM
fuelectenerg-13681	330	9			PROPN
fuelectenerg-13681	330	10	.25	.25	NUM
fuelectenerg-13681	330	11	2.35	2.35	NUM
fuelectenerg-13681	330	12			PROPN
fuelectenerg-13681	330	13	.18	.18	NUM
fuelectenerg-13681	330	14	23.56	23.56	NUM
fuelectenerg-13681	330	15			PROPN
fuelectenerg-13681	330	16	.87	.87	NUM
fuelectenerg-13681	330	17	rls	rl	NOUN
fuelectenerg-13681	330	18	0.68	0.68	NUM
fuelectenerg-13681	330	19			PROPN
fuelectenerg-13681	330	20	.05	.05	NUM
fuelectenerg-13681	330	21	12.57	12.57	NUM
fuelectenerg-13681	331	1			PROPN
fuelectenerg-13681	331	2	.96	.96	NUM
fuelectenerg-13681	331	3	4.12	4.12	NUM
fuelectenerg-13681	331	4			PROPN
fuelectenerg-13681	331	5	.32	.32	NUM
fuelectenerg-13681	331	6	87.35	87.35	NUM
fuelectenerg-13681	331	7			PROPN
fuelectenerg-13681	331	8	.54	.54	NUM
fuelectenerg-13681	331	9	mvdr	mvdr	NOUN
fuelectenerg-13681	331	10	0.75	0.75	NUM
fuelectenerg-13681	331	11			PROPN
fuelectenerg-13681	331	12	.06	.06	NUM
fuelectenerg-13681	331	13	15.32	15.32	NUM
fuelectenerg-13681	331	14			PROPN
fuelectenerg-13681	331	15	.12	.12	NUM
fuelectenerg-13681	331	16	4.58	4.58	NUM
fuelectenerg-13681	332	1			PROPN
fuelectenerg-13681	332	2	.35	.35	NUM
fuelectenerg-13681	332	3	128.47	128.47	NUM
fuelectenerg-13681	332	4			PROPN
fuelectenerg-13681	332	5	.63	.63	NUM
fuelectenerg-13681	332	6	socp	socp	NOUN
fuelectenerg-13681	332	7	2.34	2.34	NUM
fuelectenerg-13681	332	8			PROPN
fuelectenerg-13681	332	9	.18	.18	NUM
fuelectenerg-13681	332	10	48.65	48.65	NUM
fuelectenerg-13681	333	1			PROPN
fuelectenerg-13681	333	2	.67	.67	NUM
fuelectenerg-13681	333	3	14.53	14.53	NUM
fuelectenerg-13681	334	1			PROPN
fuelectenerg-13681	334	2	.15	.15	NUM
fuelectenerg-13681	334	3	312.84	312.84	NUM
fuelectenerg-13681	334	4			PROPN
fuelectenerg-13681	334	5	3.46	3.46	NUM
fuelectenerg-13681	334	6	fc	fc	PROPN
fuelectenerg-13681	334	7	-	-	PUNCT
fuelectenerg-13681	334	8	dnn	dnn	PROPN
fuelectenerg-13681	334	9	0.85	0.85	NUM
fuelectenerg-13681	335	1			PROPN
fuelectenerg-13681	335	2	.07	.07	NUM
fuelectenerg-13681	335	3	2.18	2.18	NUM
fuelectenerg-13681	335	4			PROPN
fuelectenerg-13681	335	5	.17	.17	NUM
fuelectenerg-13681	335	6	3.24	3.24	NUM
fuelectenerg-13681	335	7			PROPN
fuelectenerg-13681	335	8	.26	.26	NUM
fuelectenerg-13681	335	9	8.76	8.76	NUM
fuelectenerg-13681	335	10			PROPN
fuelectenerg-13681	335	11	.71	.71	NUM
fuelectenerg-13681	336	1	cnn	cnn	NOUN
fuelectenerg-13681	336	2	0.93	0.93	NUM
fuelectenerg-13681	336	3			PROPN
fuelectenerg-13681	336	4	.08	.08	NUM
fuelectenerg-13681	336	5	2.45	2.45	NUM
fuelectenerg-13681	336	6			PROPN
fuelectenerg-13681	336	7	.19	.19	NUM
fuelectenerg-13681	336	8	3.68	3.68	NUM
fuelectenerg-13681	336	9			PROPN
fuelectenerg-13681	336	10	.29	.29	NUM
fuelectenerg-13681	336	11	9.84	9.84	NUM
fuelectenerg-13681	336	12			PROPN
fuelectenerg-13681	336	13	.78	.78	NUM
fuelectenerg-13681	336	14	rnn	rnn	PROPN
fuelectenerg-13681	336	15	-	-	PUNCT
fuelectenerg-13681	336	16	lstm	lstm	ADJ
fuelectenerg-13681	336	17	1.12	1.12	NUM
fuelectenerg-13681	336	18			PROPN
fuelectenerg-13681	336	19	.09	.09	NUM
fuelectenerg-13681	336	20	2.87	2.87	NUM
fuelectenerg-13681	336	21			PROPN
fuelectenerg-13681	336	22	.22	.22	NUM
fuelectenerg-13681	336	23	4.53	4.53	NUM
fuelectenerg-13681	336	24			PROPN
fuelectenerg-13681	336	25	.36	.36	NUM
fuelectenerg-13681	336	26	11.32	11.32	NUM
fuelectenerg-13681	336	27			PROPN
fuelectenerg-13681	336	28	.90	.90	NUM
fuelectenerg-13681	336	29	mbdl	mbdl	NOUN
fuelectenerg-13681	336	30	0.97	0.97	NUM
fuelectenerg-13681	336	31			PROPN
fuelectenerg-13681	336	32	.08	.08	NUM
fuelectenerg-13681	336	33	2.52	2.52	NUM
fuelectenerg-13681	336	34			PROPN
fuelectenerg-13681	336	35	.20	.20	NUM
fuelectenerg-13681	336	36	3.86	3.86	NUM
fuelectenerg-13681	336	37			PROPN
fuelectenerg-13681	336	38	.31	.31	NUM
fuelectenerg-13681	336	39	10.18	10.18	NUM
fuelectenerg-13681	336	40			PROPN
fuelectenerg-13681	336	41	.82	.82	NUM
fuelectenerg-13681	336	42	the	the	DET
fuelectenerg-13681	336	43	execution	execution	NOUN
fuelectenerg-13681	336	44	time	time	NOUN
fuelectenerg-13681	336	45	measurements	measurement	NOUN
fuelectenerg-13681	336	46	on	on	ADP
fuelectenerg-13681	336	47	different	different	ADJ
fuelectenerg-13681	336	48	hardware	hardware	NOUN
fuelectenerg-13681	336	49	platforms	platform	NOUN
fuelectenerg-13681	336	50	confirm	confirm	VERB
fuelectenerg-13681	336	51	that	that	SCONJ
fuelectenerg-13681	336	52	deep	deep	ADJ
fuelectenerg-13681	336	53	learning	learning	NOUN
fuelectenerg-13681	336	54	approaches	approach	NOUN
fuelectenerg-13681	336	55	are	be	AUX
fuelectenerg-13681	336	56	particularly	particularly	ADV
fuelectenerg-13681	336	57	well	well	ADV
fuelectenerg-13681	336	58	-	-	PUNCT
fuelectenerg-13681	336	59	suited	suit	VERB
fuelectenerg-13681	336	60	for	for	ADP
fuelectenerg-13681	336	61	resource	resource	NOUN
fuelectenerg-13681	336	62	-	-	PUNCT
fuelectenerg-13681	336	63	constrained	constrain	VERB
fuelectenerg-13681	336	64	systems	system	NOUN
fuelectenerg-13681	336	65	and	and	CCONJ
fuelectenerg-13681	336	66	larger	large	ADJ
fuelectenerg-13681	336	67	mimo	mimo	PROPN
fuelectenerg-13681	336	68	configurations	configuration	NOUN
fuelectenerg-13681	336	69	.	.	PUNCT
fuelectenerg-13681	337	1	on	on	ADP
fuelectenerg-13681	337	2	the	the	DET
fuelectenerg-13681	337	3	jetson	jetson	PROPN
fuelectenerg-13681	337	4	agx	agx	PROPN
fuelectenerg-13681	337	5	xavier	xavier	PROPN
fuelectenerg-13681	337	6	,	,	PUNCT
fuelectenerg-13681	337	7	socp	socp	PROPN
fuelectenerg-13681	337	8	takes	take	VERB
fuelectenerg-13681	337	9	312.84	312.84	NUM
fuelectenerg-13681	337	10	ms	ms	NOUN
fuelectenerg-13681	337	11	for	for	ADP
fuelectenerg-13681	337	12	a	a	DET
fuelectenerg-13681	337	13	32×32	32×32	NUM
fuelectenerg-13681	337	14	mimo	mimo	NOUN
fuelectenerg-13681	337	15	system	system	NOUN
fuelectenerg-13681	337	16	,	,	PUNCT
fuelectenerg-13681	337	17	compared	compare	VERB
fuelectenerg-13681	337	18	to	to	ADP
fuelectenerg-13681	337	19	just	just	ADV
fuelectenerg-13681	337	20	9.84	9.84	NUM
fuelectenerg-13681	337	21	ms	ms	NOUN
fuelectenerg-13681	337	22	for	for	ADP
fuelectenerg-13681	337	23	cnn	cnn	PROPN
fuelectenerg-13681	337	24	–	–	PUNCT
fuelectenerg-13681	337	25	a	a	DET
fuelectenerg-13681	337	26	31.8×	31.8×	NUM
fuelectenerg-13681	337	27	speedup	speedup	NOUN
fuelectenerg-13681	337	28	(	(	PUNCT
fuelectenerg-13681	337	29	p	p	X
fuelectenerg-13681	337	30	<	<	X
fuelectenerg-13681	337	31	0.001	0.001	NUM
fuelectenerg-13681	337	32	)	)	PUNCT
fuelectenerg-13681	337	33	.	.	PUNCT
fuelectenerg-13681	338	1	it	it	PRON
fuelectenerg-13681	338	2	’s	’	VERB
fuelectenerg-13681	338	3	important	important	ADJ
fuelectenerg-13681	338	4	to	to	PART
fuelectenerg-13681	338	5	note	note	VERB
fuelectenerg-13681	338	6	that	that	SCONJ
fuelectenerg-13681	338	7	these	these	DET
fuelectenerg-13681	338	8	measurements	measurement	NOUN
fuelectenerg-13681	338	9	focus	focus	VERB
fuelectenerg-13681	338	10	on	on	ADP
fuelectenerg-13681	338	11	inference	inference	NOUN
fuelectenerg-13681	338	12	time	time	NOUN
fuelectenerg-13681	338	13	for	for	ADP
fuelectenerg-13681	338	14	deep	deep	ADJ
fuelectenerg-13681	338	15	learning	learning	NOUN
fuelectenerg-13681	338	16	approaches	approach	NOUN
fuelectenerg-13681	338	17	and	and	CCONJ
fuelectenerg-13681	338	18	do	do	AUX
fuelectenerg-13681	338	19	not	not	PART
fuelectenerg-13681	338	20	include	include	VERB
fuelectenerg-13681	338	21	the	the	DET
fuelectenerg-13681	338	22	offline	offline	ADJ
fuelectenerg-13681	338	23	training	training	NOUN
fuelectenerg-13681	338	24	time	time	NOUN
fuelectenerg-13681	338	25	.	.	PUNCT
fuelectenerg-13681	339	1	the	the	DET
fuelectenerg-13681	339	2	one	one	NUM
fuelectenerg-13681	339	3	-	-	PUNCT
fuelectenerg-13681	339	4	time	time	NOUN
fuelectenerg-13681	339	5	training	training	NOUN
fuelectenerg-13681	339	6	cost	cost	NOUN
fuelectenerg-13681	339	7	for	for	ADP
fuelectenerg-13681	339	8	the	the	DET
fuelectenerg-13681	339	9	neural	neural	ADJ
fuelectenerg-13681	339	10	networks	network	NOUN
fuelectenerg-13681	339	11	in	in	ADP
fuelectenerg-13681	339	12	our	our	PRON
fuelectenerg-13681	339	13	experiments	experiment	NOUN
fuelectenerg-13681	339	14	ranged	range	VERB
fuelectenerg-13681	339	15	from	from	ADP
fuelectenerg-13681	339	16	4.5	4.5	NUM
fuelectenerg-13681	339	17	hours	hour	NOUN
fuelectenerg-13681	339	18	(	(	PUNCT
fuelectenerg-13681	339	19	fc	fc	X
fuelectenerg-13681	339	20	-	-	PUNCT
fuelectenerg-13681	339	21	dnn	dnn	PROPN
fuelectenerg-13681	339	22	)	)	PUNCT
fuelectenerg-13681	339	23	to	to	ADP
fuelectenerg-13681	339	24	18.2	18.2	NUM
fuelectenerg-13681	339	25	hours	hour	NOUN
fuelectenerg-13681	339	26	(	(	PUNCT
fuelectenerg-13681	339	27	rnnlstm	rnnlstm	NOUN
fuelectenerg-13681	339	28	)	)	PUNCT
fuelectenerg-13681	339	29	on	on	ADP
fuelectenerg-13681	339	30	our	our	PRON
fuelectenerg-13681	339	31	gpu	gpu	NOUN
fuelectenerg-13681	339	32	-	-	PUNCT
fuelectenerg-13681	339	33	equipped	equip	VERB
fuelectenerg-13681	339	34	workstation	workstation	NOUN
fuelectenerg-13681	339	35	.	.	PUNCT
fuelectenerg-13681	340	1	memory	memory	NOUN
fuelectenerg-13681	340	2	usage	usage	NOUN
fuelectenerg-13681	340	3	analysis	analysis	NOUN
fuelectenerg-13681	340	4	shows	show	VERB
fuelectenerg-13681	340	5	that	that	SCONJ
fuelectenerg-13681	340	6	traditional	traditional	ADJ
fuelectenerg-13681	340	7	algorithms	algorithm	NOUN
fuelectenerg-13681	340	8	have	have	VERB
fuelectenerg-13681	340	9	minimal	minimal	ADJ
fuelectenerg-13681	340	10	memory	memory	NOUN
fuelectenerg-13681	340	11	footprint	footprint	NOUN
fuelectenerg-13681	340	12	for	for	ADP
fuelectenerg-13681	340	13	small	small	ADJ
fuelectenerg-13681	340	14	mimo	mimo	PROPN
fuelectenerg-13681	340	15	systems	system	NOUN
fuelectenerg-13681	340	16	but	but	CCONJ
fuelectenerg-13681	340	17	scale	scale	NOUN
fuelectenerg-13681	340	18	poorly	poorly	ADV
fuelectenerg-13681	340	19	with	with	ADP
fuelectenerg-13681	340	20	system	system	NOUN
fuelectenerg-13681	340	21	size	size	NOUN
fuelectenerg-13681	340	22	.	.	PUNCT
fuelectenerg-13681	341	1	for	for	ADP
fuelectenerg-13681	341	2	a	a	DET
fuelectenerg-13681	341	3	128×128	128×128	NUM
fuelectenerg-13681	341	4	mimo	mimo	PROPN
fuelectenerg-13681	341	5	system	system	NOUN
fuelectenerg-13681	341	6	,	,	PUNCT
fuelectenerg-13681	341	7	socp	socp	NOUN
fuelectenerg-13681	341	8	requires	require	VERB
fuelectenerg-13681	341	9	approximately	approximately	ADV
fuelectenerg-13681	341	10	524	524	NUM
fuelectenerg-13681	341	11	mb	mb	NOUN
fuelectenerg-13681	341	12	of	of	ADP
fuelectenerg-13681	341	13	memory	memory	NOUN
fuelectenerg-13681	341	14	,	,	PUNCT
fuelectenerg-13681	341	15	compared	compare	VERB
fuelectenerg-13681	341	16	to	to	ADP
fuelectenerg-13681	341	17	48	48	NUM
fuelectenerg-13681	341	18	mb	mb	NOUN
fuelectenerg-13681	341	19	for	for	ADP
fuelectenerg-13681	341	20	the	the	DET
fuelectenerg-13681	341	21	largest	large	ADJ
fuelectenerg-13681	341	22	neural	neural	ADJ
fuelectenerg-13681	341	23	network	network	NOUN
fuelectenerg-13681	341	24	(	(	PUNCT
fuelectenerg-13681	341	25	rnn	rnn	PROPN
fuelectenerg-13681	341	26	-	-	PUNCT
fuelectenerg-13681	341	27	lstm)–a	lstm)–a	NOUN
fuelectenerg-13681	341	28	reduction	reduction	NOUN
fuelectenerg-13681	341	29	of	of	ADP
fuelectenerg-13681	341	30	90.8	90.8	NUM
fuelectenerg-13681	341	31	%	%	NOUN
fuelectenerg-13681	341	32	.	.	PUNCT
fuelectenerg-13681	342	1	4.4	4.4	NUM
fuelectenerg-13681	342	2	.	.	PUNCT
fuelectenerg-13681	342	3	hybrid	hybrid	ADJ
fuelectenerg-13681	342	4	approach	approach	NOUN
fuelectenerg-13681	342	5	evaluation	evaluation	NOUN
fuelectenerg-13681	342	6	fig	fig	NOUN
fuelectenerg-13681	342	7	.	.	PUNCT
fuelectenerg-13681	343	1	4	4	NUM
fuelectenerg-13681	343	2	performance	performance	NOUN
fuelectenerg-13681	343	3	of	of	ADP
fuelectenerg-13681	343	4	the	the	DET
fuelectenerg-13681	343	5	proposed	propose	VERB
fuelectenerg-13681	343	6	hybrid	hybrid	NOUN
fuelectenerg-13681	343	7	approach	approach	NOUN
fuelectenerg-13681	343	8	across	across	ADP
fuelectenerg-13681	343	9	diverse	diverse	ADJ
fuelectenerg-13681	343	10	operating	operating	NOUN
fuelectenerg-13681	343	11	conditions	condition	NOUN
fuelectenerg-13681	343	12	compared	compare	VERB
fuelectenerg-13681	343	13	to	to	ADP
fuelectenerg-13681	343	14	individual	individual	ADJ
fuelectenerg-13681	343	15	algorithms	algorithm	NOUN
fuelectenerg-13681	343	16	.	.	PUNCT
fuelectenerg-13681	344	1	the	the	DET
fuelectenerg-13681	344	2	hybrid	hybrid	ADJ
fuelectenerg-13681	344	3	approach	approach	NOUN
fuelectenerg-13681	344	4	consistently	consistently	ADV
fuelectenerg-13681	344	5	achieves	achieve	VERB
fuelectenerg-13681	344	6	nearoptimal	nearoptimal	ADJ
fuelectenerg-13681	344	7	performance	performance	NOUN
fuelectenerg-13681	344	8	by	by	ADP
fuelectenerg-13681	344	9	dynamically	dynamically	ADV
fuelectenerg-13681	344	10	selecting	select	VERB
fuelectenerg-13681	344	11	the	the	DET
fuelectenerg-13681	344	12	most	most	ADV
fuelectenerg-13681	344	13	appropriate	appropriate	ADJ
fuelectenerg-13681	344	14	algorithm	algorithm	NOUN
fuelectenerg-13681	344	15	214	214	NUM
fuelectenerg-13681	344	16	m.	m.	NOUN
fuelectenerg-13681	344	17	mohamed	mohamed	PROPN
fuelectenerg-13681	344	18	,	,	PUNCT
fuelectenerg-13681	344	19	f.	f.	PROPN
fuelectenerg-13681	344	20	aljuaid	aljuaid	PROPN
fuelectenerg-13681	344	21	,	,	PUNCT
fuelectenerg-13681	344	22	mhd	mhd	PROPN
fuelectenerg-13681	344	23	w.	w.	PROPN
fuelectenerg-13681	344	24	koubeisi	koubeisi	VERB
fuelectenerg-13681	344	25	our	our	PRON
fuelectenerg-13681	344	26	proposed	propose	VERB
fuelectenerg-13681	344	27	hybrid	hybrid	NOUN
fuelectenerg-13681	344	28	approach	approach	NOUN
fuelectenerg-13681	344	29	dynamically	dynamically	ADV
fuelectenerg-13681	344	30	selects	select	VERB
fuelectenerg-13681	344	31	the	the	DET
fuelectenerg-13681	344	32	most	most	ADV
fuelectenerg-13681	344	33	appropriate	appropriate	ADJ
fuelectenerg-13681	344	34	algorithm	algorithm	NOUN
fuelectenerg-13681	344	35	based	base	VERB
fuelectenerg-13681	344	36	on	on	ADP
fuelectenerg-13681	344	37	the	the	DET
fuelectenerg-13681	344	38	current	current	ADJ
fuelectenerg-13681	344	39	channel	channel	NOUN
fuelectenerg-13681	344	40	and	and	CCONJ
fuelectenerg-13681	344	41	system	system	NOUN
fuelectenerg-13681	344	42	state	state	NOUN
fuelectenerg-13681	344	43	.	.	PUNCT
fuelectenerg-13681	345	1	across	across	ADP
fuelectenerg-13681	345	2	all	all	DET
fuelectenerg-13681	345	3	tested	test	VERB
fuelectenerg-13681	345	4	scenarios	scenario	NOUN
fuelectenerg-13681	345	5	,	,	PUNCT
fuelectenerg-13681	345	6	it	it	PRON
fuelectenerg-13681	345	7	achieves	achieve	VERB
fuelectenerg-13681	345	8	15.3	15.3	NUM
fuelectenerg-13681	345	9	%	%	NOUN
fuelectenerg-13681	345	10	higher	high	ADJ
fuelectenerg-13681	345	11	average	average	ADJ
fuelectenerg-13681	345	12	sinr	sinr	NOUN
fuelectenerg-13681	345	13	compared	compare	VERB
fuelectenerg-13681	345	14	to	to	ADP
fuelectenerg-13681	345	15	the	the	DET
fuelectenerg-13681	345	16	best	good	ADJ
fuelectenerg-13681	345	17	fixed	fix	VERB
fuelectenerg-13681	345	18	algorithm	algorithm	NOUN
fuelectenerg-13681	345	19	(	(	PUNCT
fuelectenerg-13681	345	20	p	p	X
fuelectenerg-13681	345	21	<	<	X
fuelectenerg-13681	345	22	0.001	0.001	NUM
fuelectenerg-13681	345	23	)	)	PUNCT
fuelectenerg-13681	345	24	.	.	PUNCT
fuelectenerg-13681	346	1	the	the	DET
fuelectenerg-13681	346	2	selection	selection	NOUN
fuelectenerg-13681	346	3	accuracy	accuracy	NOUN
fuelectenerg-13681	346	4	of	of	ADP
fuelectenerg-13681	346	5	the	the	DET
fuelectenerg-13681	346	6	hybrid	hybrid	ADJ
fuelectenerg-13681	346	7	approach	approach	NOUN
fuelectenerg-13681	346	8	improves	improve	VERB
fuelectenerg-13681	346	9	over	over	ADP
fuelectenerg-13681	346	10	time	time	NOUN
fuelectenerg-13681	346	11	as	as	SCONJ
fuelectenerg-13681	346	12	it	it	PRON
fuelectenerg-13681	346	13	learns	learn	VERB
fuelectenerg-13681	346	14	from	from	ADP
fuelectenerg-13681	346	15	observed	observed	ADJ
fuelectenerg-13681	346	16	performance	performance	NOUN
fuelectenerg-13681	346	17	.	.	PUNCT
fuelectenerg-13681	347	1	after	after	ADP
fuelectenerg-13681	347	2	approximately	approximately	ADV
fuelectenerg-13681	347	3	1,000	1,000	NUM
fuelectenerg-13681	347	4	iterations	iteration	NOUN
fuelectenerg-13681	347	5	,	,	PUNCT
fuelectenerg-13681	347	6	the	the	DET
fuelectenerg-13681	347	7	selection	selection	NOUN
fuelectenerg-13681	347	8	accuracy	accuracy	NOUN
fuelectenerg-13681	347	9	stabilizes	stabilize	VERB
fuelectenerg-13681	347	10	at	at	ADP
fuelectenerg-13681	347	11	92.7	92.7	NUM
fuelectenerg-13681	347	12	%	%	NOUN
fuelectenerg-13681	347	13	,	,	PUNCT
fuelectenerg-13681	347	14	meaning	mean	VERB
fuelectenerg-13681	347	15	that	that	SCONJ
fuelectenerg-13681	347	16	it	it	PRON
fuelectenerg-13681	347	17	selects	select	VERB
fuelectenerg-13681	347	18	the	the	DET
fuelectenerg-13681	347	19	optimal	optimal	ADJ
fuelectenerg-13681	347	20	algorithm	algorithm	NOUN
fuelectenerg-13681	347	21	for	for	ADP
fuelectenerg-13681	347	22	the	the	DET
fuelectenerg-13681	347	23	current	current	ADJ
fuelectenerg-13681	347	24	conditions	condition	NOUN
fuelectenerg-13681	347	25	92.7	92.7	NUM
fuelectenerg-13681	347	26	%	%	NOUN
fuelectenerg-13681	347	27	of	of	ADP
fuelectenerg-13681	347	28	the	the	DET
fuelectenerg-13681	347	29	time	time	NOUN
fuelectenerg-13681	347	30	.	.	PUNCT
fuelectenerg-13681	348	1	the	the	DET
fuelectenerg-13681	348	2	computational	computational	ADJ
fuelectenerg-13681	348	3	overhead	overhead	NOUN
fuelectenerg-13681	348	4	of	of	ADP
fuelectenerg-13681	348	5	the	the	DET
fuelectenerg-13681	348	6	classifier	classifier	NOUN
fuelectenerg-13681	348	7	is	be	AUX
fuelectenerg-13681	348	8	negligible	negligible	ADJ
fuelectenerg-13681	348	9	compared	compare	VERB
fuelectenerg-13681	348	10	to	to	ADP
fuelectenerg-13681	348	11	the	the	DET
fuelectenerg-13681	348	12	performance	performance	NOUN
fuelectenerg-13681	348	13	gains	gain	VERB
fuelectenerg-13681	348	14	it	it	PRON
fuelectenerg-13681	348	15	provides	provide	VERB
fuelectenerg-13681	348	16	.	.	PUNCT
fuelectenerg-13681	349	1	the	the	DET
fuelectenerg-13681	349	2	selection	selection	NOUN
fuelectenerg-13681	349	3	mechanism	mechanism	NOUN
fuelectenerg-13681	349	4	requires	require	VERB
fuelectenerg-13681	349	5	approximately	approximately	ADV
fuelectenerg-13681	349	6	0.12	0.12	NUM
fuelectenerg-13681	349	7	ms	ms	NOUN
fuelectenerg-13681	349	8	per	per	ADP
fuelectenerg-13681	349	9	decision	decision	NOUN
fuelectenerg-13681	349	10	,	,	PUNCT
fuelectenerg-13681	349	11	which	which	PRON
fuelectenerg-13681	349	12	is	be	AUX
fuelectenerg-13681	349	13	less	less	ADJ
fuelectenerg-13681	349	14	than	than	ADP
fuelectenerg-13681	349	15	5	5	NUM
fuelectenerg-13681	349	16	%	%	NOUN
fuelectenerg-13681	349	17	of	of	ADP
fuelectenerg-13681	349	18	the	the	DET
fuelectenerg-13681	349	19	execution	execution	NOUN
fuelectenerg-13681	349	20	time	time	NOUN
fuelectenerg-13681	349	21	of	of	ADP
fuelectenerg-13681	349	22	even	even	ADV
fuelectenerg-13681	349	23	the	the	DET
fuelectenerg-13681	349	24	fastest	fast	ADJ
fuelectenerg-13681	349	25	beamforming	beamforme	VERB
fuelectenerg-13681	349	26	algorithm	algorithm	NOUN
fuelectenerg-13681	349	27	.	.	PUNCT
fuelectenerg-13681	350	1	the	the	DET
fuelectenerg-13681	350	2	algorithm	algorithm	NOUN
fuelectenerg-13681	350	3	selection	selection	NOUN
fuelectenerg-13681	350	4	pattern	pattern	NOUN
fuelectenerg-13681	350	5	of	of	ADP
fuelectenerg-13681	350	6	the	the	DET
fuelectenerg-13681	350	7	hybrid	hybrid	ADJ
fuelectenerg-13681	350	8	approach	approach	NOUN
fuelectenerg-13681	350	9	varies	vary	VERB
fuelectenerg-13681	350	10	across	across	ADP
fuelectenerg-13681	350	11	different	different	ADJ
fuelectenerg-13681	350	12	mobility	mobility	NOUN
fuelectenerg-13681	350	13	scenarios	scenario	NOUN
fuelectenerg-13681	350	14	.	.	PUNCT
fuelectenerg-13681	351	1	in	in	ADP
fuelectenerg-13681	351	2	static	static	ADJ
fuelectenerg-13681	351	3	conditions	condition	NOUN
fuelectenerg-13681	351	4	,	,	PUNCT
fuelectenerg-13681	351	5	it	it	PRON
fuelectenerg-13681	351	6	predominantly	predominantly	ADV
fuelectenerg-13681	351	7	selects	select	VERB
fuelectenerg-13681	351	8	traditional	traditional	ADJ
fuelectenerg-13681	351	9	algorithms	algorithm	NOUN
fuelectenerg-13681	351	10	(	(	PUNCT
fuelectenerg-13681	351	11	mvdr	mvdr	NOUN
fuelectenerg-13681	351	12	,	,	PUNCT
fuelectenerg-13681	351	13	socp	socp	ADJ
fuelectenerg-13681	351	14	)	)	PUNCT
fuelectenerg-13681	351	15	,	,	PUNCT
fuelectenerg-13681	351	16	while	while	SCONJ
fuelectenerg-13681	351	17	in	in	ADP
fuelectenerg-13681	351	18	high	high	ADJ
fuelectenerg-13681	351	19	-	-	PUNCT
fuelectenerg-13681	351	20	mobility	mobility	NOUN
fuelectenerg-13681	351	21	scenarios	scenario	NOUN
fuelectenerg-13681	351	22	,	,	PUNCT
fuelectenerg-13681	351	23	it	it	PRON
fuelectenerg-13681	351	24	favors	favor	VERB
fuelectenerg-13681	351	25	deep	deep	ADJ
fuelectenerg-13681	351	26	learning	learning	NOUN
fuelectenerg-13681	351	27	approaches	approach	NOUN
fuelectenerg-13681	351	28	(	(	PUNCT
fuelectenerg-13681	351	29	rnn	rnn	NOUN
fuelectenerg-13681	351	30	-	-	PUNCT
fuelectenerg-13681	351	31	lstm	lstm	PROPN
fuelectenerg-13681	351	32	,	,	PUNCT
fuelectenerg-13681	351	33	cnn	cnn	PROPN
fuelectenerg-13681	351	34	)	)	PUNCT
fuelectenerg-13681	351	35	.	.	PUNCT
fuelectenerg-13681	352	1	we	we	PRON
fuelectenerg-13681	352	2	analyzed	analyze	VERB
fuelectenerg-13681	352	3	the	the	DET
fuelectenerg-13681	352	4	regret	regret	NOUN
fuelectenerg-13681	352	5	of	of	ADP
fuelectenerg-13681	352	6	the	the	DET
fuelectenerg-13681	352	7	hybrid	hybrid	ADJ
fuelectenerg-13681	352	8	approach	approach	NOUN
fuelectenerg-13681	352	9	,	,	PUNCT
fuelectenerg-13681	352	10	defined	define	VERB
fuelectenerg-13681	352	11	as	as	ADP
fuelectenerg-13681	352	12	the	the	DET
fuelectenerg-13681	352	13	cumulative	cumulative	ADJ
fuelectenerg-13681	352	14	difference	difference	NOUN
fuelectenerg-13681	352	15	between	between	ADP
fuelectenerg-13681	352	16	the	the	DET
fuelectenerg-13681	352	17	performance	performance	NOUN
fuelectenerg-13681	352	18	of	of	ADP
fuelectenerg-13681	352	19	the	the	DET
fuelectenerg-13681	352	20	hybrid	hybrid	ADJ
fuelectenerg-13681	352	21	approach	approach	NOUN
fuelectenerg-13681	352	22	and	and	CCONJ
fuelectenerg-13681	352	23	that	that	PRON
fuelectenerg-13681	352	24	of	of	ADP
fuelectenerg-13681	352	25	the	the	DET
fuelectenerg-13681	352	26	optimal	optimal	ADJ
fuelectenerg-13681	352	27	algorithm	algorithm	NOUN
fuelectenerg-13681	352	28	for	for	ADP
fuelectenerg-13681	352	29	each	each	DET
fuelectenerg-13681	352	30	condition	condition	NOUN
fuelectenerg-13681	352	31	.	.	PUNCT
fuelectenerg-13681	353	1	the	the	DET
fuelectenerg-13681	353	2	empirical	empirical	ADJ
fuelectenerg-13681	353	3	regret	regret	NOUN
fuelectenerg-13681	353	4	grows	grow	VERB
fuelectenerg-13681	353	5	sub	sub	ADJ
fuelectenerg-13681	353	6	-	-	ADJ
fuelectenerg-13681	353	7	linearly	linearly	ADV
fuelectenerg-13681	353	8	with	with	ADP
fuelectenerg-13681	353	9	time	time	NOUN
fuelectenerg-13681	353	10	,	,	PUNCT
fuelectenerg-13681	353	11	confirming	confirm	VERB
fuelectenerg-13681	353	12	our	our	PRON
fuelectenerg-13681	353	13	theoretical	theoretical	ADJ
fuelectenerg-13681	353	14	bound	bind	VERB
fuelectenerg-13681	353	15	of	of	ADP
fuelectenerg-13681	353	16	(	(	PUNCT
fuelectenerg-13681	353	17	log	log	PROPN
fuelectenerg-13681	353	18	)	)	PUNCT
fuelectenerg-13681	353	19	o	o	NOUN
fuelectenerg-13681	353	20	t	t	PROPN
fuelectenerg-13681	353	21	t	t	PROPN
fuelectenerg-13681	353	22	.	.	PUNCT
fuelectenerg-13681	354	1	the	the	DET
fuelectenerg-13681	354	2	hybrid	hybrid	ADJ
fuelectenerg-13681	354	3	approach	approach	NOUN
fuelectenerg-13681	354	4	also	also	ADV
fuelectenerg-13681	354	5	exhibits	exhibit	VERB
fuelectenerg-13681	354	6	excellent	excellent	ADJ
fuelectenerg-13681	354	7	adaptability	adaptability	NOUN
fuelectenerg-13681	354	8	to	to	ADP
fuelectenerg-13681	354	9	non	non	ADJ
fuelectenerg-13681	354	10	-	-	ADJ
fuelectenerg-13681	354	11	stationary	stationary	ADJ
fuelectenerg-13681	354	12	envinments	envinment	NOUN
fuelectenerg-13681	354	13	.	.	PUNCT
fuelectenerg-13681	355	1	when	when	SCONJ
fuelectenerg-13681	355	2	we	we	PRON
fuelectenerg-13681	355	3	introduced	introduce	VERB
fuelectenerg-13681	355	4	abrupt	abrupt	ADJ
fuelectenerg-13681	355	5	changes	change	NOUN
fuelectenerg-13681	355	6	in	in	ADP
fuelectenerg-13681	355	7	channel	channel	NOUN
fuelectenerg-13681	355	8	conditions	condition	NOUN
fuelectenerg-13681	355	9	,	,	PUNCT
fuelectenerg-13681	355	10	the	the	DET
fuelectenerg-13681	355	11	approach	approach	NOUN
fuelectenerg-13681	355	12	quickly	quickly	ADV
fuelectenerg-13681	355	13	adapted	adapt	VERB
fuelectenerg-13681	355	14	its	its	PRON
fuelectenerg-13681	355	15	selection	selection	NOUN
fuelectenerg-13681	355	16	strategy	strategy	NOUN
fuelectenerg-13681	355	17	,	,	PUNCT
fuelectenerg-13681	355	18	typically	typically	ADV
fuelectenerg-13681	355	19	within	within	ADP
fuelectenerg-13681	355	20	5	5	NUM
fuelectenerg-13681	355	21	-	-	SYM
fuelectenerg-13681	355	22	10	10	NUM
fuelectenerg-13681	355	23	iterations	iteration	NOUN
fuelectenerg-13681	355	24	.	.	PUNCT
fuelectenerg-13681	356	1	4.5	4.5	NUM
fuelectenerg-13681	356	2	.	.	PUNCT
fuelectenerg-13681	356	3	hardware	hardware	NOUN
fuelectenerg-13681	356	4	validation	validation	NOUN
fuelectenerg-13681	356	5	and	and	CCONJ
fuelectenerg-13681	356	6	real	real	ADJ
fuelectenerg-13681	356	7	-	-	PUNCT
fuelectenerg-13681	356	8	world	world	NOUN
fuelectenerg-13681	356	9	testing	testing	NOUN
fuelectenerg-13681	356	10	to	to	PART
fuelectenerg-13681	356	11	validate	validate	VERB
fuelectenerg-13681	356	12	our	our	PRON
fuelectenerg-13681	356	13	simulation	simulation	NOUN
fuelectenerg-13681	356	14	results	result	NOUN
fuelectenerg-13681	356	15	in	in	ADP
fuelectenerg-13681	356	16	real	real	ADJ
fuelectenerg-13681	356	17	-	-	PUNCT
fuelectenerg-13681	356	18	world	world	NOUN
fuelectenerg-13681	356	19	conditions	condition	NOUN
fuelectenerg-13681	357	1	,	,	PUNCT
fuelectenerg-13681	357	2	we	we	PRON
fuelectenerg-13681	357	3	implemented	implement	VERB
fuelectenerg-13681	357	4	key	key	ADJ
fuelectenerg-13681	357	5	algorithms	algorithm	NOUN
fuelectenerg-13681	357	6	on	on	ADP
fuelectenerg-13681	357	7	usrp	usrp	NOUN
fuelectenerg-13681	357	8	x310	x310	PROPN
fuelectenerg-13681	357	9	software	software	NOUN
fuelectenerg-13681	357	10	-	-	PUNCT
fuelectenerg-13681	357	11	defined	define	VERB
fuelectenerg-13681	357	12	radio	radio	NOUN
fuelectenerg-13681	357	13	platforms	platform	NOUN
fuelectenerg-13681	357	14	.	.	PUNCT
fuelectenerg-13681	358	1	the	the	DET
fuelectenerg-13681	358	2	hardware	hardware	NOUN
fuelectenerg-13681	358	3	testbed	testbe	VERB
fuelectenerg-13681	358	4	consisted	consist	VERB
fuelectenerg-13681	358	5	of	of	ADP
fuelectenerg-13681	358	6	2	2	NUM
fuelectenerg-13681	358	7	usrp	usrp	NOUN
fuelectenerg-13681	358	8	x310	x310	PROPN
fuelectenerg-13681	358	9	devices	device	NOUN
fuelectenerg-13681	358	10	with	with	ADP
fuelectenerg-13681	358	11	ubx-160	ubx-160	PROPN
fuelectenerg-13681	358	12	daughterboards	daughterboard	NOUN
fuelectenerg-13681	358	13	(	(	PUNCT
fuelectenerg-13681	358	14	one	one	NUM
fuelectenerg-13681	358	15	transmitter	transmitter	NOUN
fuelectenerg-13681	358	16	,	,	PUNCT
fuelectenerg-13681	358	17	one	one	NUM
fuelectenerg-13681	358	18	receiver	receiver	NOUN
fuelectenerg-13681	358	19	)	)	PUNCT
fuelectenerg-13681	358	20	,	,	PUNCT
fuelectenerg-13681	358	21	4	4	NUM
fuelectenerg-13681	358	22	omnidirectional	omnidirectional	ADJ
fuelectenerg-13681	358	23	antennas	antenna	NOUN
fuelectenerg-13681	358	24	per	per	ADP
fuelectenerg-13681	358	25	device	device	NOUN
fuelectenerg-13681	358	26	(	(	PUNCT
fuelectenerg-13681	358	27	4×4	4×4	NUM
fuelectenerg-13681	358	28	mimo	mimo	PROPN
fuelectenerg-13681	358	29	configuration	configuration	PROPN
fuelectenerg-13681	358	30	)	)	PUNCT
fuelectenerg-13681	358	31	,	,	PUNCT
fuelectenerg-13681	358	32	gnu	gnu	X
fuelectenerg-13681	358	33	radio	radio	NOUN
fuelectenerg-13681	358	34	software	software	NOUN
fuelectenerg-13681	358	35	framework	framework	NOUN
fuelectenerg-13681	358	36	for	for	ADP
fuelectenerg-13681	358	37	signal	signal	NOUN
fuelectenerg-13681	358	38	processing	processing	NOUN
fuelectenerg-13681	358	39	,	,	PUNCT
fuelectenerg-13681	358	40	and	and	CCONJ
fuelectenerg-13681	358	41	a	a	DET
fuelectenerg-13681	358	42	channel	channel	NOUN
fuelectenerg-13681	358	43	emulator	emulator	NOUN
fuelectenerg-13681	358	44	for	for	ADP
fuelectenerg-13681	358	45	controlled	control	VERB
fuelectenerg-13681	358	46	testing	testing	NOUN
fuelectenerg-13681	358	47	.	.	PUNCT
fuelectenerg-13681	359	1	the	the	DET
fuelectenerg-13681	359	2	hardware	hardware	NOUN
fuelectenerg-13681	359	3	experiments	experiment	NOUN
fuelectenerg-13681	359	4	confirmed	confirm	VERB
fuelectenerg-13681	359	5	our	our	PRON
fuelectenerg-13681	359	6	simulation	simulation	NOUN
fuelectenerg-13681	359	7	results	result	NOUN
fuelectenerg-13681	359	8	with	with	ADP
fuelectenerg-13681	359	9	high	high	ADJ
fuelectenerg-13681	359	10	fidelity	fidelity	NOUN
fuelectenerg-13681	359	11	.	.	PUNCT
fuelectenerg-13681	360	1	the	the	DET
fuelectenerg-13681	360	2	relative	relative	ADJ
fuelectenerg-13681	360	3	performance	performance	NOUN
fuelectenerg-13681	360	4	ranking	ranking	NOUN
fuelectenerg-13681	360	5	of	of	ADP
fuelectenerg-13681	360	6	algorithms	algorithm	NOUN
fuelectenerg-13681	360	7	remained	remain	VERB
fuelectenerg-13681	360	8	consistent	consistent	ADJ
fuelectenerg-13681	360	9	across	across	ADP
fuelectenerg-13681	360	10	simulation	simulation	NOUN
fuelectenerg-13681	360	11	and	and	CCONJ
fuelectenerg-13681	360	12	hardware	hardware	NOUN
fuelectenerg-13681	360	13	implementation	implementation	NOUN
fuelectenerg-13681	360	14	,	,	PUNCT
fuelectenerg-13681	360	15	with	with	ADP
fuelectenerg-13681	360	16	correlation	correlation	NOUN
fuelectenerg-13681	360	17	coefficient	coefficient	NOUN
fuelectenerg-13681	360	18	r	r	NOUN
fuelectenerg-13681	360	19	=	=	SYM
fuelectenerg-13681	360	20	0.94	0.94	NUM
fuelectenerg-13681	360	21	(	(	PUNCT
fuelectenerg-13681	360	22	p	p	X
fuelectenerg-13681	360	23	<	<	X
fuelectenerg-13681	360	24	0.001	0.001	NUM
fuelectenerg-13681	360	25	)	)	PUNCT
fuelectenerg-13681	360	26	between	between	ADP
fuelectenerg-13681	360	27	simulated	simulate	VERB
fuelectenerg-13681	360	28	and	and	CCONJ
fuelectenerg-13681	360	29	measured	measure	VERB
fuelectenerg-13681	360	30	sinr	sinr	NOUN
fuelectenerg-13681	360	31	values	value	NOUN
fuelectenerg-13681	360	32	.	.	PUNCT
fuelectenerg-13681	361	1	in	in	ADP
fuelectenerg-13681	361	2	an	an	DET
fuelectenerg-13681	361	3	office	office	NOUN
fuelectenerg-13681	361	4	environment	environment	NOUN
fuelectenerg-13681	361	5	with	with	ADP
fuelectenerg-13681	361	6	moderate	moderate	ADJ
fuelectenerg-13681	361	7	mobility	mobility	NOUN
fuelectenerg-13681	361	8	,	,	PUNCT
fuelectenerg-13681	361	9	we	we	PRON
fuelectenerg-13681	361	10	measured	measure	VERB
fuelectenerg-13681	361	11	an	an	DET
fuelectenerg-13681	361	12	average	average	ADJ
fuelectenerg-13681	361	13	sinr	sinr	NOUN
fuelectenerg-13681	361	14	of	of	ADP
fuelectenerg-13681	361	15	18.3	18.3	NUM
fuelectenerg-13681	361	16	db	db	NOUN
fuelectenerg-13681	361	17	for	for	ADP
fuelectenerg-13681	361	18	mvdr	mvdr	NOUN
fuelectenerg-13681	361	19	,	,	PUNCT
fuelectenerg-13681	361	20	17.8	17.8	NUM
fuelectenerg-13681	361	21	db	db	NOUN
fuelectenerg-13681	361	22	for	for	ADP
fuelectenerg-13681	361	23	rls	rls	PROPN
fuelectenerg-13681	361	24	,	,	PUNCT
fuelectenerg-13681	361	25	19.5	19.5	NUM
fuelectenerg-13681	361	26	db	db	PROPN
fuelectenerg-13681	361	27	for	for	ADP
fuelectenerg-13681	361	28	cnn	cnn	PROPN
fuelectenerg-13681	361	29	,	,	PUNCT
fuelectenerg-13681	361	30	and	and	CCONJ
fuelectenerg-13681	361	31	20.1	20.1	NUM
fuelectenerg-13681	361	32	db	db	NOUN
fuelectenerg-13681	361	33	for	for	ADP
fuelectenerg-13681	361	34	rnn	rnn	NOUN
fuelectenerg-13681	361	35	-	-	PUNCT
fuelectenerg-13681	361	36	lstm	lstm	NOUN
fuelectenerg-13681	361	37	.	.	PUNCT
fuelectenerg-13681	362	1	these	these	DET
fuelectenerg-13681	362	2	values	value	NOUN
fuelectenerg-13681	362	3	were	be	AUX
fuelectenerg-13681	362	4	within	within	ADP
fuelectenerg-13681	362	5	1.2	1.2	NUM
fuelectenerg-13681	362	6	db	db	PROPN
fuelectenerg-13681	362	7	of	of	ADP
fuelectenerg-13681	362	8	our	our	PRON
fuelectenerg-13681	362	9	simulation	simulation	NOUN
fuelectenerg-13681	362	10	predictions	prediction	NOUN
fuelectenerg-13681	362	11	.	.	PUNCT
fuelectenerg-13681	363	1	the	the	DET
fuelectenerg-13681	363	2	computational	computational	ADJ
fuelectenerg-13681	363	3	efficiency	efficiency	NOUN
fuelectenerg-13681	363	4	measurements	measurement	NOUN
fuelectenerg-13681	363	5	on	on	ADP
fuelectenerg-13681	363	6	the	the	DET
fuelectenerg-13681	363	7	embedded	embed	VERB
fuelectenerg-13681	363	8	platform	platform	NOUN
fuelectenerg-13681	363	9	(	(	PUNCT
fuelectenerg-13681	363	10	usrp	usrp	NOUN
fuelectenerg-13681	363	11	onboard	onboard	ADP
fuelectenerg-13681	363	12	cpu	cpu	PROPN
fuelectenerg-13681	363	13	)	)	PUNCT
fuelectenerg-13681	363	14	matched	match	VERB
fuelectenerg-13681	363	15	our	our	PRON
fuelectenerg-13681	363	16	jetson	jetson	PROPN
fuelectenerg-13681	363	17	agx	agx	PROPN
fuelectenerg-13681	363	18	xavier	xavier	PROPN
fuelectenerg-13681	363	19	results	result	VERB
fuelectenerg-13681	363	20	within	within	ADP
fuelectenerg-13681	363	21	a	a	DET
fuelectenerg-13681	363	22	margin	margin	NOUN
fuelectenerg-13681	363	23	of	of	ADP
fuelectenerg-13681	363	24	15	15	NUM
fuelectenerg-13681	363	25	%	%	NOUN
fuelectenerg-13681	363	26	,	,	PUNCT
fuelectenerg-13681	363	27	confirming	confirm	VERB
fuelectenerg-13681	363	28	the	the	DET
fuelectenerg-13681	363	29	superior	superior	ADJ
fuelectenerg-13681	363	30	scalability	scalability	NOUN
fuelectenerg-13681	363	31	of	of	ADP
fuelectenerg-13681	363	32	deep	deep	ADJ
fuelectenerg-13681	363	33	learning	learning	NOUN
fuelectenerg-13681	363	34	approaches	approach	NOUN
fuelectenerg-13681	363	35	for	for	ADP
fuelectenerg-13681	363	36	resource	resource	NOUN
fuelectenerg-13681	363	37	-	-	PUNCT
fuelectenerg-13681	363	38	constrained	constrain	VERB
fuelectenerg-13681	363	39	implementations	implementation	NOUN
fuelectenerg-13681	363	40	.	.	PUNCT
fuelectenerg-13681	364	1	4.6	4.6	NUM
fuelectenerg-13681	364	2	.	.	PUNCT
fuelectenerg-13681	364	3	discussion	discussion	NOUN
fuelectenerg-13681	364	4	our	our	PRON
fuelectenerg-13681	364	5	comprehensive	comprehensive	ADJ
fuelectenerg-13681	364	6	analysis	analysis	NOUN
fuelectenerg-13681	364	7	reveals	reveal	VERB
fuelectenerg-13681	364	8	that	that	SCONJ
fuelectenerg-13681	364	9	the	the	DET
fuelectenerg-13681	364	10	choice	choice	NOUN
fuelectenerg-13681	364	11	between	between	ADP
fuelectenerg-13681	364	12	traditional	traditional	ADJ
fuelectenerg-13681	364	13	optimization	optimization	NOUN
fuelectenerg-13681	364	14	algorithms	algorithm	NOUN
fuelectenerg-13681	364	15	and	and	CCONJ
fuelectenerg-13681	364	16	deep	deep	ADJ
fuelectenerg-13681	364	17	learning	learning	NOUN
fuelectenerg-13681	364	18	approaches	approach	NOUN
fuelectenerg-13681	364	19	for	for	ADP
fuelectenerg-13681	364	20	adaptive	adaptive	ADJ
fuelectenerg-13681	364	21	beamforming	beamforming	NOUN
fuelectenerg-13681	364	22	depends	depend	VERB
fuelectenerg-13681	364	23	critically	critically	ADV
fuelectenerg-13681	364	24	on	on	ADP
fuelectenerg-13681	364	25	the	the	DET
fuelectenerg-13681	364	26	specific	specific	ADJ
fuelectenerg-13681	364	27	operating	operating	NOUN
fuelectenerg-13681	364	28	conditions	condition	NOUN
fuelectenerg-13681	364	29	and	and	CCONJ
fuelectenerg-13681	364	30	system	system	NOUN
fuelectenerg-13681	364	31	requirements	requirement	NOUN
fuelectenerg-13681	364	32	.	.	PUNCT
fuelectenerg-13681	365	1	traditional	traditional	ADJ
fuelectenerg-13681	365	2	algorithms	algorithm	NOUN
fuelectenerg-13681	365	3	are	be	AUX
fuelectenerg-13681	365	4	preferable	preferable	ADJ
fuelectenerg-13681	365	5	in	in	ADP
fuelectenerg-13681	365	6	steady	steady	ADJ
fuelectenerg-13681	365	7	-	-	PUNCT
fuelectenerg-13681	365	8	state	state	NOUN
fuelectenerg-13681	365	9	,	,	PUNCT
fuelectenerg-13681	365	10	slowly	slowly	ADV
fuelectenerg-13681	365	11	varying	vary	VERB
fuelectenerg-13681	365	12	channels	channel	NOUN
fuelectenerg-13681	365	13	where	where	SCONJ
fuelectenerg-13681	365	14	convergence	convergence	NOUN
fuelectenerg-13681	365	15	time	time	NOUN
fuelectenerg-13681	365	16	is	be	AUX
fuelectenerg-13681	365	17	not	not	PART
fuelectenerg-13681	365	18	critical	critical	ADJ
fuelectenerg-13681	365	19	;	;	PUNCT
fuelectenerg-13681	365	20	low	low	ADJ
fuelectenerg-13681	365	21	-	-	PUNCT
fuelectenerg-13681	365	22	snr	snr	NOUN
fuelectenerg-13681	365	23	environments	environment	NOUN
fuelectenerg-13681	365	24	where	where	SCONJ
fuelectenerg-13681	365	25	statistical	statistical	ADJ
fuelectenerg-13681	365	26	optimality	optimality	NOUN
fuelectenerg-13681	365	27	is	be	AUX
fuelectenerg-13681	365	28	crucial	crucial	ADJ
fuelectenerg-13681	365	29	;	;	PUNCT
fuelectenerg-13681	365	30	small	small	ADJ
fuelectenerg-13681	365	31	-	-	PUNCT
fuelectenerg-13681	365	32	scale	scale	NOUN
fuelectenerg-13681	365	33	mimo	mimo	PROPN
fuelectenerg-13681	365	34	systems	system	NOUN
fuelectenerg-13681	365	35	where	where	SCONJ
fuelectenerg-13681	365	36	computational	computational	ADJ
fuelectenerg-13681	365	37	complexity	complexity	NOUN
fuelectenerg-13681	365	38	is	be	AUX
fuelectenerg-13681	365	39	manageable	manageable	ADJ
fuelectenerg-13681	365	40	;	;	PUNCT
fuelectenerg-13681	365	41	and	and	CCONJ
fuelectenerg-13681	365	42	scenarios	scenario	NOUN
fuelectenerg-13681	365	43	requiring	require	VERB
fuelectenerg-13681	365	44	theoretical	theoretical	ADJ
fuelectenerg-13681	365	45	performance	performance	NOUN
fuelectenerg-13681	365	46	guarantees	guarantee	NOUN
fuelectenerg-13681	365	47	.	.	PUNCT
fuelectenerg-13681	366	1	deep	deep	ADJ
fuelectenerg-13681	366	2	learning	learning	NOUN
fuelectenerg-13681	366	3	approaches	approach	NOUN
fuelectenerg-13681	366	4	are	be	AUX
fuelectenerg-13681	366	5	advantageous	advantageous	ADJ
fuelectenerg-13681	366	6	in	in	ADP
fuelectenerg-13681	366	7	dynamic	dynamic	ADJ
fuelectenerg-13681	366	8	,	,	PUNCT
fuelectenerg-13681	366	9	rapidly	rapidly	ADV
fuelectenerg-13681	366	10	varying	vary	VERB
fuelectenerg-13681	366	11	channels	channel	NOUN
fuelectenerg-13681	366	12	requiring	require	VERB
fuelectenerg-13681	366	13	fast	fast	ADJ
fuelectenerg-13681	366	14	adaptation	adaptation	NOUN
fuelectenerg-13681	366	15	;	;	PUNCT
fuelectenerg-13681	366	16	complex	complex	ADJ
fuelectenerg-13681	366	17	propagation	propagation	NOUN
fuelectenerg-13681	366	18	environments	environment	NOUN
fuelectenerg-13681	366	19	with	with	ADP
fuelectenerg-13681	366	20	rich	rich	ADJ
fuelectenerg-13681	366	21	deep	deep	ADJ
fuelectenerg-13681	366	22	learning	learning	NOUN
fuelectenerg-13681	366	23	and	and	CCONJ
fuelectenerg-13681	366	24	traditional	traditional	ADJ
fuelectenerg-13681	366	25	algorithms	algorithm	NOUN
fuelectenerg-13681	366	26	for	for	ADP
fuelectenerg-13681	366	27	mimo	mimo	PROPN
fuelectenerg-13681	366	28	beamforming	beamforme	VERB
fuelectenerg-13681	366	29	215	215	NUM
fuelectenerg-13681	366	30	multipath	multipath	NOUN
fuelectenerg-13681	366	31	;	;	PUNCT
fuelectenerg-13681	366	32	resource	resource	NOUN
fuelectenerg-13681	366	33	-	-	PUNCT
fuelectenerg-13681	366	34	constrained	constrain	VERB
fuelectenerg-13681	366	35	or	or	CCONJ
fuelectenerg-13681	366	36	latency	latency	NOUN
fuelectenerg-13681	366	37	-	-	PUNCT
fuelectenerg-13681	366	38	sensitive	sensitive	ADJ
fuelectenerg-13681	366	39	implementations	implementation	NOUN
fuelectenerg-13681	366	40	;	;	PUNCT
fuelectenerg-13681	366	41	and	and	CCONJ
fuelectenerg-13681	366	42	massive	massive	ADJ
fuelectenerg-13681	366	43	mimo	mimo	PROPN
fuelectenerg-13681	366	44	systems	system	NOUN
fuelectenerg-13681	366	45	where	where	SCONJ
fuelectenerg-13681	366	46	traditional	traditional	ADJ
fuelectenerg-13681	366	47	methods	method	NOUN
fuelectenerg-13681	366	48	become	become	VERB
fuelectenerg-13681	366	49	computationally	computationally	ADV
fuelectenerg-13681	366	50	prohibitive	prohibitive	ADJ
fuelectenerg-13681	366	51	.	.	PUNCT
fuelectenerg-13681	367	1	the	the	DET
fuelectenerg-13681	367	2	proposed	propose	VERB
fuelectenerg-13681	367	3	hybrid	hybrid	NOUN
fuelectenerg-13681	367	4	approach	approach	NOUN
fuelectenerg-13681	367	5	successfully	successfully	ADV
fuelectenerg-13681	367	6	leverages	leverage	VERB
fuelectenerg-13681	367	7	the	the	DET
fuelectenerg-13681	367	8	complementary	complementary	ADJ
fuelectenerg-13681	367	9	strengths	strength	NOUN
fuelectenerg-13681	367	10	of	of	ADP
fuelectenerg-13681	367	11	both	both	DET
fuelectenerg-13681	367	12	methodologies	methodology	NOUN
fuelectenerg-13681	367	13	,	,	PUNCT
fuelectenerg-13681	367	14	achieving	achieve	VERB
fuelectenerg-13681	367	15	consistently	consistently	ADV
fuelectenerg-13681	367	16	superior	superior	ADJ
fuelectenerg-13681	367	17	performance	performance	NOUN
fuelectenerg-13681	367	18	across	across	ADP
fuelectenerg-13681	367	19	diverse	diverse	ADJ
fuelectenerg-13681	367	20	operating	operating	NOUN
fuelectenerg-13681	367	21	conditions	condition	NOUN
fuelectenerg-13681	367	22	.	.	PUNCT
fuelectenerg-13681	368	1	its	its	PRON
fuelectenerg-13681	368	2	ability	ability	NOUN
fuelectenerg-13681	368	3	to	to	PART
fuelectenerg-13681	368	4	dynamically	dynamically	VERB
fuelectenerg-13681	368	5	select	select	VERB
fuelectenerg-13681	368	6	the	the	DET
fuelectenerg-13681	368	7	most	most	ADV
fuelectenerg-13681	368	8	appropriate	appropriate	ADJ
fuelectenerg-13681	368	9	algorithm	algorithm	NOUN
fuelectenerg-13681	368	10	based	base	VERB
fuelectenerg-13681	368	11	on	on	ADP
fuelectenerg-13681	368	12	current	current	ADJ
fuelectenerg-13681	368	13	conditions	condition	NOUN
fuelectenerg-13681	368	14	makes	make	VERB
fuelectenerg-13681	368	15	it	it	PRON
fuelectenerg-13681	368	16	particularly	particularly	ADV
fuelectenerg-13681	368	17	well	well	ADV
fuelectenerg-13681	368	18	-	-	PUNCT
fuelectenerg-13681	368	19	suited	suit	VERB
fuelectenerg-13681	368	20	for	for	ADP
fuelectenerg-13681	368	21	heterogeneous	heterogeneous	ADJ
fuelectenerg-13681	368	22	deployment	deployment	NOUN
fuelectenerg-13681	368	23	scenarios	scenario	NOUN
fuelectenerg-13681	368	24	.	.	PUNCT
fuelectenerg-13681	369	1	from	from	ADP
fuelectenerg-13681	369	2	a	a	DET
fuelectenerg-13681	369	3	theoretical	theoretical	ADJ
fuelectenerg-13681	369	4	perspective	perspective	NOUN
fuelectenerg-13681	369	5	,	,	PUNCT
fuelectenerg-13681	369	6	the	the	DET
fuelectenerg-13681	369	7	performance	performance	NOUN
fuelectenerg-13681	369	8	differences	difference	NOUN
fuelectenerg-13681	369	9	between	between	ADP
fuelectenerg-13681	369	10	traditional	traditional	ADJ
fuelectenerg-13681	369	11	and	and	CCONJ
fuelectenerg-13681	369	12	deep	deep	ADJ
fuelectenerg-13681	369	13	learning	learning	NOUN
fuelectenerg-13681	369	14	approaches	approach	NOUN
fuelectenerg-13681	369	15	can	can	AUX
fuelectenerg-13681	369	16	be	be	AUX
fuelectenerg-13681	369	17	understood	understand	VERB
fuelectenerg-13681	369	18	in	in	ADP
fuelectenerg-13681	369	19	terms	term	NOUN
fuelectenerg-13681	369	20	of	of	ADP
fuelectenerg-13681	369	21	the	the	DET
fuelectenerg-13681	369	22	bias	bias	NOUN
fuelectenerg-13681	369	23	-	-	PUNCT
fuelectenerg-13681	369	24	variance	variance	NOUN
fuelectenerg-13681	369	25	tradeoff	tradeoff	NOUN
fuelectenerg-13681	369	26	.	.	PUNCT
fuelectenerg-13681	370	1	traditional	traditional	ADJ
fuelectenerg-13681	370	2	algorithms	algorithm	NOUN
fuelectenerg-13681	370	3	have	have	VERB
fuelectenerg-13681	370	4	lower	low	ADJ
fuelectenerg-13681	370	5	bias	bias	NOUN
fuelectenerg-13681	370	6	(	(	PUNCT
fuelectenerg-13681	370	7	they	they	PRON
fuelectenerg-13681	370	8	approach	approach	VERB
fuelectenerg-13681	370	9	optimal	optimal	ADJ
fuelectenerg-13681	370	10	performance	performance	NOUN
fuelectenerg-13681	370	11	given	give	VERB
fuelectenerg-13681	370	12	accurate	accurate	ADJ
fuelectenerg-13681	370	13	models	model	NOUN
fuelectenerg-13681	370	14	and	and	CCONJ
fuelectenerg-13681	370	15	sufficient	sufficient	ADJ
fuelectenerg-13681	370	16	iterations	iteration	NOUN
fuelectenerg-13681	370	17	)	)	PUNCT
fuelectenerg-13681	370	18	but	but	CCONJ
fuelectenerg-13681	370	19	higher	high	ADJ
fuelectenerg-13681	370	20	variance	variance	NOUN
fuelectenerg-13681	370	21	(	(	PUNCT
fuelectenerg-13681	370	22	sensitivity	sensitivity	NOUN
fuelectenerg-13681	370	23	to	to	ADP
fuelectenerg-13681	370	24	model	model	NOUN
fuelectenerg-13681	370	25	mismatch	mismatch	PROPN
fuelectenerg-13681	370	26	)	)	PUNCT
fuelectenerg-13681	370	27	.	.	PUNCT
fuelectenerg-13681	371	1	deep	deep	ADJ
fuelectenerg-13681	371	2	learning	learning	NOUN
fuelectenerg-13681	371	3	approaches	approach	NOUN
fuelectenerg-13681	371	4	have	have	VERB
fuelectenerg-13681	371	5	higher	high	ADJ
fuelectenerg-13681	371	6	bias	bias	NOUN
fuelectenerg-13681	371	7	(	(	PUNCT
fuelectenerg-13681	371	8	they	they	PRON
fuelectenerg-13681	371	9	may	may	AUX
fuelectenerg-13681	371	10	not	not	PART
fuelectenerg-13681	371	11	achieve	achieve	VERB
fuelectenerg-13681	371	12	theoretical	theoretical	ADJ
fuelectenerg-13681	371	13	optimality	optimality	NOUN
fuelectenerg-13681	371	14	)	)	PUNCT
fuelectenerg-13681	371	15	but	but	CCONJ
fuelectenerg-13681	371	16	lower	low	ADJ
fuelectenerg-13681	371	17	variance	variance	NOUN
fuelectenerg-13681	371	18	(	(	PUNCT
fuelectenerg-13681	371	19	greater	great	ADJ
fuelectenerg-13681	371	20	robustness	robustness	NOUN
fuelectenerg-13681	371	21	to	to	ADP
fuelectenerg-13681	371	22	model	model	NOUN
fuelectenerg-13681	371	23	mismatch	mismatch	PROPN
fuelectenerg-13681	371	24	)	)	PUNCT
fuelectenerg-13681	371	25	.	.	PUNCT
fuelectenerg-13681	372	1	it	it	PRON
fuelectenerg-13681	372	2	’s	’	VERB
fuelectenerg-13681	372	3	important	important	ADJ
fuelectenerg-13681	372	4	to	to	PART
fuelectenerg-13681	372	5	acknowledge	acknowledge	VERB
fuelectenerg-13681	372	6	the	the	DET
fuelectenerg-13681	372	7	limitations	limitation	NOUN
fuelectenerg-13681	372	8	of	of	ADP
fuelectenerg-13681	372	9	our	our	PRON
fuelectenerg-13681	372	10	study	study	NOUN
fuelectenerg-13681	372	11	.	.	PUNCT
fuelectenerg-13681	373	1	deep	deep	ADJ
fuelectenerg-13681	373	2	learning	learning	NOUN
fuelectenerg-13681	373	3	approaches	approach	NOUN
fuelectenerg-13681	373	4	require	require	VERB
fuelectenerg-13681	373	5	significant	significant	ADJ
fuelectenerg-13681	373	6	offline	offline	ADJ
fuelectenerg-13681	373	7	training	training	NOUN
fuelectenerg-13681	373	8	data	datum	NOUN
fuelectenerg-13681	373	9	and	and	CCONJ
fuelectenerg-13681	373	10	computational	computational	ADJ
fuelectenerg-13681	373	11	resources	resource	NOUN
fuelectenerg-13681	373	12	.	.	PUNCT
fuelectenerg-13681	374	1	while	while	SCONJ
fuelectenerg-13681	374	2	inference	inference	NOUN
fuelectenerg-13681	374	3	is	be	AUX
fuelectenerg-13681	374	4	efficient	efficient	ADJ
fuelectenerg-13681	374	5	,	,	PUNCT
fuelectenerg-13681	374	6	the	the	DET
fuelectenerg-13681	374	7	one	one	NUM
fuelectenerg-13681	374	8	-	-	PUNCT
fuelectenerg-13681	374	9	time	time	NOUN
fuelectenerg-13681	374	10	training	training	NOUN
fuelectenerg-13681	374	11	cost	cost	NOUN
fuelectenerg-13681	374	12	can	can	AUX
fuelectenerg-13681	374	13	be	be	AUX
fuelectenerg-13681	374	14	substantial	substantial	ADJ
fuelectenerg-13681	374	15	.	.	PUNCT
fuelectenerg-13681	375	1	additionally	additionally	ADV
fuelectenerg-13681	375	2	,	,	PUNCT
fuelectenerg-13681	375	3	neural	neural	ADJ
fuelectenerg-13681	375	4	networks	network	NOUN
fuelectenerg-13681	375	5	lack	lack	VERB
fuelectenerg-13681	375	6	the	the	DET
fuelectenerg-13681	375	7	theoretical	theoretical	ADJ
fuelectenerg-13681	375	8	guarantees	guarantee	NOUN
fuelectenerg-13681	375	9	of	of	ADP
fuelectenerg-13681	375	10	traditional	traditional	ADJ
fuelectenerg-13681	375	11	methods	method	NOUN
fuelectenerg-13681	375	12	,	,	PUNCT
fuelectenerg-13681	375	13	which	which	PRON
fuelectenerg-13681	375	14	may	may	AUX
fuelectenerg-13681	375	15	be	be	AUX
fuelectenerg-13681	375	16	problematic	problematic	ADJ
fuelectenerg-13681	375	17	for	for	ADP
fuelectenerg-13681	375	18	safety	safety	NOUN
fuelectenerg-13681	375	19	-	-	PUNCT
fuelectenerg-13681	375	20	critical	critical	ADJ
fuelectenerg-13681	375	21	applications	application	NOUN
fuelectenerg-13681	375	22	.	.	PUNCT
fuelectenerg-13681	376	1	the	the	DET
fuelectenerg-13681	376	2	superiority	superiority	NOUN
fuelectenerg-13681	376	3	of	of	ADP
fuelectenerg-13681	376	4	model	model	NOUN
fuelectenerg-13681	376	5	-	-	PUNCT
fuelectenerg-13681	376	6	based	base	VERB
fuelectenerg-13681	376	7	deep	deep	ADJ
fuelectenerg-13681	376	8	learning	learning	NOUN
fuelectenerg-13681	376	9	(	(	PUNCT
fuelectenerg-13681	376	10	mbdl	mbdl	PROPN
fuelectenerg-13681	376	11	)	)	PUNCT
fuelectenerg-13681	376	12	over	over	ADP
fuelectenerg-13681	376	13	pure	pure	ADJ
fuelectenerg-13681	376	14	data	data	NOUN
fuelectenerg-13681	376	15	-	-	PUNCT
fuelectenerg-13681	376	16	driven	drive	VERB
fuelectenerg-13681	376	17	approaches	approach	NOUN
fuelectenerg-13681	376	18	highlights	highlight	VERB
fuelectenerg-13681	376	19	the	the	DET
fuelectenerg-13681	376	20	importance	importance	NOUN
fuelectenerg-13681	376	21	of	of	ADP
fuelectenerg-13681	376	22	incorporating	incorporate	VERB
fuelectenerg-13681	376	23	domain	domain	NOUN
fuelectenerg-13681	376	24	knowledge	knowledge	NOUN
fuelectenerg-13681	376	25	into	into	ADP
fuelectenerg-13681	376	26	neural	neural	ADJ
fuelectenerg-13681	376	27	network	network	NOUN
fuelectenerg-13681	376	28	design	design	NOUN
fuelectenerg-13681	376	29	.	.	PUNCT
fuelectenerg-13681	377	1	by	by	ADP
fuelectenerg-13681	377	2	combining	combine	VERB
fuelectenerg-13681	377	3	the	the	DET
fuelectenerg-13681	377	4	analytical	analytical	ADJ
fuelectenerg-13681	377	5	structure	structure	NOUN
fuelectenerg-13681	377	6	of	of	ADP
fuelectenerg-13681	377	7	traditional	traditional	ADJ
fuelectenerg-13681	377	8	algorithms	algorithm	NOUN
fuelectenerg-13681	377	9	with	with	ADP
fuelectenerg-13681	377	10	the	the	DET
fuelectenerg-13681	377	11	adaptability	adaptability	NOUN
fuelectenerg-13681	377	12	of	of	ADP
fuelectenerg-13681	377	13	deep	deep	ADJ
fuelectenerg-13681	377	14	learning	learning	NOUN
fuelectenerg-13681	377	15	,	,	PUNCT
fuelectenerg-13681	377	16	mbdl	mbdl	PROPN
fuelectenerg-13681	377	17	achieves	achieve	VERB
fuelectenerg-13681	377	18	a	a	DET
fuelectenerg-13681	377	19	favorable	favorable	ADJ
fuelectenerg-13681	377	20	balance	balance	NOUN
fuelectenerg-13681	377	21	between	between	ADP
fuelectenerg-13681	377	22	performance	performance	NOUN
fuelectenerg-13681	377	23	,	,	PUNCT
fuelectenerg-13681	377	24	efficiency	efficiency	NOUN
fuelectenerg-13681	377	25	,	,	PUNCT
fuelectenerg-13681	377	26	and	and	CCONJ
fuelectenerg-13681	377	27	interpretability	interpretability	NOUN
fuelectenerg-13681	377	28	.	.	PUNCT
fuelectenerg-13681	378	1	5	5	X
fuelectenerg-13681	378	2	.	.	X
fuelectenerg-13681	378	3	conclusion	conclusion	NOUN
fuelectenerg-13681	378	4	this	this	DET
fuelectenerg-13681	378	5	paper	paper	NOUN
fuelectenerg-13681	378	6	presented	present	VERB
fuelectenerg-13681	378	7	a	a	DET
fuelectenerg-13681	378	8	comprehensive	comprehensive	ADJ
fuelectenerg-13681	378	9	comparative	comparative	ADJ
fuelectenerg-13681	378	10	analysis	analysis	NOUN
fuelectenerg-13681	378	11	of	of	ADP
fuelectenerg-13681	378	12	deep	deep	ADJ
fuelectenerg-13681	378	13	learning	learning	NOUN
fuelectenerg-13681	378	14	approaches	approach	NOUN
fuelectenerg-13681	378	15	and	and	CCONJ
fuelectenerg-13681	378	16	traditional	traditional	ADJ
fuelectenerg-13681	378	17	optimization	optimization	NOUN
fuelectenerg-13681	378	18	algorithms	algorithm	NOUN
fuelectenerg-13681	378	19	for	for	ADP
fuelectenerg-13681	378	20	adaptive	adaptive	ADJ
fuelectenerg-13681	378	21	beamforming	beamforming	NOUN
fuelectenerg-13681	378	22	in	in	ADP
fuelectenerg-13681	378	23	mimo	mimo	PROPN
fuelectenerg-13681	378	24	systems	system	NOUN
fuelectenerg-13681	378	25	.	.	PUNCT
fuelectenerg-13681	379	1	through	through	ADP
fuelectenerg-13681	379	2	extensive	extensive	ADJ
fuelectenerg-13681	379	3	evaluation	evaluation	NOUN
fuelectenerg-13681	379	4	across	across	ADP
fuelectenerg-13681	379	5	diverse	diverse	ADJ
fuelectenerg-13681	379	6	operating	operating	NOUN
fuelectenerg-13681	379	7	conditions	condition	NOUN
fuelectenerg-13681	379	8	,	,	PUNCT
fuelectenerg-13681	379	9	we	we	PRON
fuelectenerg-13681	379	10	identified	identify	VERB
fuelectenerg-13681	379	11	the	the	DET
fuelectenerg-13681	379	12	specific	specific	ADJ
fuelectenerg-13681	379	13	strengths	strength	NOUN
fuelectenerg-13681	379	14	,	,	PUNCT
fuelectenerg-13681	379	15	limitations	limitation	NOUN
fuelectenerg-13681	379	16	,	,	PUNCT
fuelectenerg-13681	379	17	and	and	CCONJ
fuelectenerg-13681	379	18	operational	operational	ADJ
fuelectenerg-13681	379	19	regimes	regime	NOUN
fuelectenerg-13681	379	20	of	of	ADP
fuelectenerg-13681	379	21	each	each	DET
fuelectenerg-13681	379	22	methodology	methodology	NOUN
fuelectenerg-13681	379	23	.	.	PUNCT
fuelectenerg-13681	380	1	traditional	traditional	ADJ
fuelectenerg-13681	380	2	optimization	optimization	NOUN
fuelectenerg-13681	380	3	algorithms	algorithm	NOUN
fuelectenerg-13681	380	4	achieve	achieve	VERB
fuelectenerg-13681	380	5	superior	superior	ADJ
fuelectenerg-13681	380	6	performance	performance	NOUN
fuelectenerg-13681	380	7	in	in	ADP
fuelectenerg-13681	380	8	static	static	ADJ
fuelectenerg-13681	380	9	environments	environment	NOUN
fuelectenerg-13681	380	10	,	,	PUNCT
fuelectenerg-13681	380	11	particularly	particularly	ADV
fuelectenerg-13681	380	12	in	in	ADP
fuelectenerg-13681	380	13	low	low	ADJ
fuelectenerg-13681	380	14	-	-	PUNCT
fuelectenerg-13681	380	15	snr	snr	NOUN
fuelectenerg-13681	380	16	regimes	regime	NOUN
fuelectenerg-13681	380	17	where	where	SCONJ
fuelectenerg-13681	380	18	they	they	PRON
fuelectenerg-13681	380	19	outperform	outperform	VERB
fuelectenerg-13681	380	20	deep	deep	ADJ
fuelectenerg-13681	380	21	learning	learning	NOUN
fuelectenerg-13681	380	22	approaches	approach	NOUN
fuelectenerg-13681	380	23	by	by	ADP
fuelectenerg-13681	380	24	1.2–2.8	1.2–2.8	NUM
fuelectenerg-13681	380	25	db	db	NOUN
fuelectenerg-13681	380	26	in	in	ADP
fuelectenerg-13681	380	27	sinr	sinr	NOUN
fuelectenerg-13681	380	28	(	(	PUNCT
fuelectenerg-13681	380	29	p	p	X
fuelectenerg-13681	380	30	<	<	X
fuelectenerg-13681	380	31	0.01	0.01	NUM
fuelectenerg-13681	380	32	)	)	PUNCT
fuelectenerg-13681	380	33	.	.	PUNCT
fuelectenerg-13681	381	1	deep	deep	ADJ
fuelectenerg-13681	381	2	learning	learning	NOUN
fuelectenerg-13681	381	3	approaches	approach	NOUN
fuelectenerg-13681	381	4	demonstrate	demonstrate	VERB
fuelectenerg-13681	381	5	advantages	advantage	NOUN
fuelectenerg-13681	381	6	in	in	ADP
fuelectenerg-13681	381	7	dynamic	dynamic	ADJ
fuelectenerg-13681	381	8	environments	environment	NOUN
fuelectenerg-13681	381	9	,	,	PUNCT
fuelectenerg-13681	381	10	achieving	achieve	VERB
fuelectenerg-13681	381	11	23.7	23.7	NUM
fuelectenerg-13681	381	12	%	%	NOUN
fuelectenerg-13681	381	13	higher	high	ADJ
fuelectenerg-13681	381	14	average	average	ADJ
fuelectenerg-13681	381	15	sinr	sinr	NOUN
fuelectenerg-13681	381	16	(	(	PUNCT
fuelectenerg-13681	381	17	p	p	X
fuelectenerg-13681	381	18	<	<	X
fuelectenerg-13681	381	19	0.01	0.01	NUM
fuelectenerg-13681	381	20	)	)	PUNCT
fuelectenerg-13681	381	21	in	in	ADP
fuelectenerg-13681	381	22	high	high	ADJ
fuelectenerg-13681	381	23	-	-	PUNCT
fuelectenerg-13681	381	24	mobility	mobility	NOUN
fuelectenerg-13681	381	25	scenarios	scenario	NOUN
fuelectenerg-13681	381	26	and	and	CCONJ
fuelectenerg-13681	381	27	convergence	convergence	NOUN
fuelectenerg-13681	381	28	times	time	NOUN
fuelectenerg-13681	381	29	6	6	NUM
fuelectenerg-13681	381	30	–	–	SYM
fuelectenerg-13681	381	31	13	13	NUM
fuelectenerg-13681	381	32	times	time	NOUN
fuelectenerg-13681	381	33	faster	fast	ADV
fuelectenerg-13681	381	34	than	than	ADP
fuelectenerg-13681	381	35	traditional	traditional	ADJ
fuelectenerg-13681	381	36	methods	method	NOUN
fuelectenerg-13681	381	37	.	.	PUNCT
fuelectenerg-13681	382	1	computational	computational	ADJ
fuelectenerg-13681	382	2	complexity	complexity	NOUN
fuelectenerg-13681	382	3	analysis	analysis	NOUN
fuelectenerg-13681	382	4	reveals	reveal	VERB
fuelectenerg-13681	382	5	that	that	SCONJ
fuelectenerg-13681	382	6	traditional	traditional	ADJ
fuelectenerg-13681	382	7	methods	method	NOUN
fuelectenerg-13681	382	8	scale	scale	VERB
fuelectenerg-13681	382	9	as	as	ADP
fuelectenerg-13681	382	10	o(n³	o(n³	PROPN
fuelectenerg-13681	382	11	)	)	PUNCT
fuelectenerg-13681	382	12	with	with	ADP
fuelectenerg-13681	382	13	mimo	mimo	PROPN
fuelectenerg-13681	382	14	size	size	NOUN
fuelectenerg-13681	382	15	n	n	CCONJ
fuelectenerg-13681	382	16	,	,	PUNCT
fuelectenerg-13681	382	17	while	while	SCONJ
fuelectenerg-13681	382	18	deep	deep	ADJ
fuelectenerg-13681	382	19	learning	learning	NOUN
fuelectenerg-13681	382	20	maintains	maintain	VERB
fuelectenerg-13681	382	21	o(n	o(n	NUM
fuelectenerg-13681	382	22	)	)	PUNCT
fuelectenerg-13681	382	23	inference	inference	NOUN
fuelectenerg-13681	382	24	complexity	complexity	NOUN
fuelectenerg-13681	382	25	.	.	PUNCT
fuelectenerg-13681	383	1	our	our	PRON
fuelectenerg-13681	383	2	proposed	propose	VERB
fuelectenerg-13681	383	3	hybrid	hybrid	NOUN
fuelectenerg-13681	383	4	approach	approach	NOUN
fuelectenerg-13681	383	5	leverages	leverage	VERB
fuelectenerg-13681	383	6	the	the	DET
fuelectenerg-13681	383	7	complementary	complementary	ADJ
fuelectenerg-13681	383	8	strengths	strength	NOUN
fuelectenerg-13681	383	9	of	of	ADP
fuelectenerg-13681	383	10	both	both	DET
fuelectenerg-13681	383	11	methodologies	methodology	NOUN
fuelectenerg-13681	383	12	,	,	PUNCT
fuelectenerg-13681	383	13	achieving	achieve	VERB
fuelectenerg-13681	383	14	15.3	15.3	NUM
fuelectenerg-13681	383	15	%	%	NOUN
fuelectenerg-13681	383	16	higher	high	ADJ
fuelectenerg-13681	383	17	average	average	ADJ
fuelectenerg-13681	383	18	sinr	sinr	NOUN
fuelectenerg-13681	383	19	in	in	ADP
fuelectenerg-13681	383	20	mixed	mixed	ADJ
fuelectenerg-13681	383	21	scenarios	scenario	NOUN
fuelectenerg-13681	383	22	compared	compare	VERB
fuelectenerg-13681	383	23	to	to	ADP
fuelectenerg-13681	383	24	fixed	fix	VERB
fuelectenerg-13681	383	25	algorithms	algorithm	NOUN
fuelectenerg-13681	383	26	.	.	PUNCT
fuelectenerg-13681	384	1	future	future	ADJ
fuelectenerg-13681	384	2	research	research	NOUN
fuelectenerg-13681	384	3	should	should	AUX
fuelectenerg-13681	384	4	focus	focus	VERB
fuelectenerg-13681	384	5	on	on	ADP
fuelectenerg-13681	384	6	developing	develop	VERB
fuelectenerg-13681	384	7	mathematically	mathematically	ADV
fuelectenerg-13681	384	8	rigorous	rigorous	ADJ
fuelectenerg-13681	384	9	foundations	foundation	NOUN
fuelectenerg-13681	384	10	for	for	ADP
fuelectenerg-13681	384	11	deep	deep	ADJ
fuelectenerg-13681	384	12	learning	learning	NOUN
fuelectenerg-13681	384	13	approaches	approach	NOUN
fuelectenerg-13681	384	14	in	in	ADP
fuelectenerg-13681	384	15	wireless	wireless	ADJ
fuelectenerg-13681	384	16	communications	communication	NOUN
fuelectenerg-13681	384	17	,	,	PUNCT
fuelectenerg-13681	384	18	creating	create	VERB
fuelectenerg-13681	384	19	more	more	ADV
fuelectenerg-13681	384	20	sophisticated	sophisticated	ADJ
fuelectenerg-13681	384	21	hybrid	hybrid	ADJ
fuelectenerg-13681	384	22	frameworks	framework	NOUN
fuelectenerg-13681	384	23	,	,	PUNCT
fuelectenerg-13681	384	24	exploring	explore	VERB
fuelectenerg-13681	384	25	lifelong	lifelong	ADJ
fuelectenerg-13681	384	26	learning	learning	NOUN
fuelectenerg-13681	384	27	approaches	approach	NOUN
fuelectenerg-13681	384	28	,	,	PUNCT
fuelectenerg-13681	384	29	extending	extend	VERB
fuelectenerg-13681	384	30	the	the	DET
fuelectenerg-13681	384	31	analysis	analysis	NOUN
fuelectenerg-13681	384	32	to	to	ADP
fuelectenerg-13681	384	33	emerging	emerge	VERB
fuelectenerg-13681	384	34	technologies	technology	NOUN
fuelectenerg-13681	384	35	such	such	ADJ
fuelectenerg-13681	384	36	as	as	ADP
fuelectenerg-13681	384	37	reconfigurable	reconfigurable	ADJ
fuelectenerg-13681	384	38	intelligent	intelligent	ADJ
fuelectenerg-13681	384	39	surfaces	surface	NOUN
fuelectenerg-13681	384	40	,	,	PUNCT
fuelectenerg-13681	384	41	and	and	CCONJ
fuelectenerg-13681	384	42	investigating	investigate	VERB
fuelectenerg-13681	384	43	hardwareaware	hardwareaware	ADJ
fuelectenerg-13681	384	44	neural	neural	ADJ
fuelectenerg-13681	384	45	network	network	NOUN
fuelectenerg-13681	384	46	designs	design	NOUN
fuelectenerg-13681	384	47	.	.	PUNCT
fuelectenerg-13681	385	1	while	while	SCONJ
fuelectenerg-13681	385	2	neither	neither	CCONJ
fuelectenerg-13681	385	3	traditional	traditional	ADJ
fuelectenerg-13681	385	4	optimization	optimization	NOUN
fuelectenerg-13681	385	5	algorithms	algorithm	NOUN
fuelectenerg-13681	385	6	nor	nor	CCONJ
fuelectenerg-13681	385	7	deep	deep	ADJ
fuelectenerg-13681	385	8	learning	learning	NOUN
fuelectenerg-13681	385	9	approaches	approach	VERB
fuelectenerg-13681	385	10	uniformly	uniformly	ADV
fuelectenerg-13681	385	11	dominate	dominate	VERB
fuelectenerg-13681	385	12	across	across	ADP
fuelectenerg-13681	385	13	all	all	DET
fuelectenerg-13681	385	14	scenarios	scenario	NOUN
fuelectenerg-13681	385	15	,	,	PUNCT
fuelectenerg-13681	385	16	understanding	understand	VERB
fuelectenerg-13681	385	17	their	their	PRON
fuelectenerg-13681	385	18	relative	relative	ADJ
fuelectenerg-13681	385	19	strengths	strength	NOUN
fuelectenerg-13681	385	20	enables	enable	VERB
fuelectenerg-13681	385	21	the	the	DET
fuelectenerg-13681	385	22	development	development	NOUN
fuelectenerg-13681	385	23	of	of	ADP
fuelectenerg-13681	385	24	adaptive	adaptive	ADJ
fuelectenerg-13681	385	25	systems	system	NOUN
fuelectenerg-13681	385	26	that	that	PRON
fuelectenerg-13681	385	27	leverage	leverage	VERB
fuelectenerg-13681	385	28	the	the	DET
fuelectenerg-13681	385	29	most	most	ADV
fuelectenerg-13681	385	30	appropriate	appropriate	ADJ
fuelectenerg-13681	385	31	technique	technique	NOUN
fuelectenerg-13681	385	32	for	for	ADP
fuelectenerg-13681	385	33	each	each	DET
fuelectenerg-13681	385	34	specific	specific	ADJ
fuelectenerg-13681	385	35	condition	condition	NOUN
fuelectenerg-13681	385	36	.	.	PUNCT
fuelectenerg-13681	386	1	216	216	NUM
fuelectenerg-13681	386	2	m.	m.	NOUN
fuelectenerg-13681	386	3	mohamed	mohamed	PROPN
fuelectenerg-13681	386	4	,	,	PUNCT
fuelectenerg-13681	386	5	f.	f.	PROPN
fuelectenerg-13681	386	6	aljuaid	aljuaid	PROPN
fuelectenerg-13681	386	7	,	,	PUNCT
fuelectenerg-13681	386	8	mhd	mhd	PROPN
fuelectenerg-13681	386	9	w.	w.	PROPN
fuelectenerg-13681	386	10	koubeisi	koubeisi	PROPN
fuelectenerg-13681	386	11	references	reference	NOUN
fuelectenerg-13681	386	12	[	[	X
fuelectenerg-13681	386	13	1	1	NUM
fuelectenerg-13681	386	14	]	]	PUNCT
fuelectenerg-13681	386	15	c.	c.	PROPN
fuelectenerg-13681	386	16	liaskos	liaskos	PROPN
fuelectenerg-13681	386	17	,	,	PUNCT
fuelectenerg-13681	386	18	s.	s.	PROPN
fuelectenerg-13681	386	19	nie	nie	PROPN
fuelectenerg-13681	386	20	,	,	PUNCT
fuelectenerg-13681	386	21	a.	a.	NOUN
fuelectenerg-13681	386	22	tsioliaridou	tsioliaridou	PROPN
fuelectenerg-13681	386	23	,	,	PUNCT
fuelectenerg-13681	386	24	a.	a.	NOUN
fuelectenerg-13681	386	25	pitsillides	pitsillides	PROPN
fuelectenerg-13681	386	26	,	,	PUNCT
fuelectenerg-13681	386	27	s.	s.	PROPN
fuelectenerg-13681	386	28	ioannidis	ioannidis	PROPN
fuelectenerg-13681	386	29	and	and	CCONJ
fuelectenerg-13681	386	30	i.	i.	PROPN
fuelectenerg-13681	386	31	akyildiz	akyildiz	PROPN
fuelectenerg-13681	386	32	,	,	PUNCT
fuelectenerg-13681	386	33	"	"	PUNCT
fuelectenerg-13681	386	34	a	a	DET
fuelectenerg-13681	386	35	new	new	ADJ
fuelectenerg-13681	386	36	wireless	wireless	ADJ
fuelectenerg-13681	386	37	communication	communication	NOUN
fuelectenerg-13681	386	38	paradigm	paradigm	NOUN
fuelectenerg-13681	386	39	through	through	ADP
fuelectenerg-13681	386	40	software	software	NOUN
fuelectenerg-13681	386	41	-	-	PUNCT
fuelectenerg-13681	386	42	controlled	control	VERB
fuelectenerg-13681	386	43	metasurfaces	metasurface	NOUN
fuelectenerg-13681	386	44	,	,	PUNCT
fuelectenerg-13681	386	45	"	"	PUNCT
fuelectenerg-13681	386	46	ieee	ieee	X
fuelectenerg-13681	386	47	commun	commun	PROPN
fuelectenerg-13681	386	48	.	.	PUNCT
fuelectenerg-13681	387	1	mag	mag	PROPN
fuelectenerg-13681	387	2	.	.	PROPN
fuelectenerg-13681	387	3	,	,	PUNCT
fuelectenerg-13681	387	4	vol	vol	NOUN
fuelectenerg-13681	387	5	.	.	PROPN
fuelectenerg-13681	388	1	56	56	NUM
fuelectenerg-13681	388	2	,	,	PUNCT
fuelectenerg-13681	388	3	no	no	INTJ
fuelectenerg-13681	388	4	.	.	NOUN
fuelectenerg-13681	388	5	9	9	NUM
fuelectenerg-13681	388	6	,	,	PUNCT
fuelectenerg-13681	388	7	pp	pp	ADJ
fuelectenerg-13681	388	8	.	.	PUNCT
fuelectenerg-13681	389	1	162–169	162–169	NUM
fuelectenerg-13681	389	2	,	,	PUNCT
fuelectenerg-13681	389	3	2018	2018	NUM
fuelectenerg-13681	389	4	.	.	PUNCT
fuelectenerg-13681	390	1	[	[	X
fuelectenerg-13681	390	2	2	2	NUM
fuelectenerg-13681	390	3	]	]	PUNCT
fuelectenerg-13681	390	4	m.	m.	NOUN
fuelectenerg-13681	390	5	z.	z.	PROPN
fuelectenerg-13681	390	6	chowdhury	chowdhury	PROPN
fuelectenerg-13681	390	7	,	,	PUNCT
fuelectenerg-13681	390	8	m.	m.	NOUN
fuelectenerg-13681	390	9	shahjalal	shahjalal	PROPN
fuelectenerg-13681	390	10	,	,	PUNCT
fuelectenerg-13681	390	11	s.	s.	PROPN
fuelectenerg-13681	390	12	ahmed	ahmed	PROPN
fuelectenerg-13681	390	13	and	and	CCONJ
fuelectenerg-13681	390	14	y.	y.	PROPN
fuelectenerg-13681	390	15	m.	m.	PROPN
fuelectenerg-13681	390	16	jang	jang	PROPN
fuelectenerg-13681	390	17	,	,	PUNCT
fuelectenerg-13681	390	18	"	"	PUNCT
fuelectenerg-13681	390	19	6	6	NUM
fuelectenerg-13681	390	20	g	g	NOUN
fuelectenerg-13681	390	21	wireless	wireless	ADJ
fuelectenerg-13681	390	22	communication	communication	NOUN
fuelectenerg-13681	390	23	systems	system	NOUN
fuelectenerg-13681	390	24	:	:	PUNCT
fuelectenerg-13681	390	25	applications	application	NOUN
fuelectenerg-13681	390	26	,	,	PUNCT
fuelectenerg-13681	390	27	requirements	requirement	NOUN
fuelectenerg-13681	390	28	,	,	PUNCT
fuelectenerg-13681	390	29	technologies	technology	NOUN
fuelectenerg-13681	390	30	,	,	PUNCT
fuelectenerg-13681	390	31	challenges	challenge	NOUN
fuelectenerg-13681	390	32	,	,	PUNCT
fuelectenerg-13681	390	33	and	and	CCONJ
fuelectenerg-13681	390	34	research	research	NOUN
fuelectenerg-13681	390	35	directions	direction	NOUN
fuelectenerg-13681	390	36	,	,	PUNCT
fuelectenerg-13681	390	37	"	"	PUNCT
fuelectenerg-13681	390	38	ieee	ieee	PROPN
fuelectenerg-13681	390	39	open	open	PROPN
fuelectenerg-13681	390	40	j.	j.	PROPN
fuelectenerg-13681	390	41	commun	commun	PROPN
fuelectenerg-13681	390	42	.	.	PUNCT
fuelectenerg-13681	391	1	soc	soc	PROPN
fuelectenerg-13681	391	2	.	.	PUNCT
fuelectenerg-13681	391	3	,	,	PUNCT
fuelectenerg-13681	391	4	vol	vol	NOUN
fuelectenerg-13681	391	5	.	.	PROPN
fuelectenerg-13681	391	6	1	1	NUM
fuelectenerg-13681	391	7	,	,	PUNCT
fuelectenerg-13681	391	8	pp	pp	ADJ
fuelectenerg-13681	391	9	.	.	PUNCT
fuelectenerg-13681	392	1	957–975	957–975	NUM
fuelectenerg-13681	392	2	,	,	PUNCT
fuelectenerg-13681	392	3	2020	2020	NUM
fuelectenerg-13681	392	4	.	.	PUNCT
fuelectenerg-13681	393	1	[	[	X
fuelectenerg-13681	393	2	3	3	X
fuelectenerg-13681	393	3	]	]	PUNCT
fuelectenerg-13681	393	4	w.	w.	PROPN
fuelectenerg-13681	393	5	saad	saad	PROPN
fuelectenerg-13681	393	6	,	,	PUNCT
fuelectenerg-13681	393	7	m.	m.	NOUN
fuelectenerg-13681	393	8	bennis	bennis	PROPN
fuelectenerg-13681	393	9	and	and	CCONJ
fuelectenerg-13681	393	10	m.	m.	PROPN
fuelectenerg-13681	393	11	chen	chen	PROPN
fuelectenerg-13681	393	12	,	,	PUNCT
fuelectenerg-13681	393	13	"	"	PUNCT
fuelectenerg-13681	393	14	a	a	DET
fuelectenerg-13681	393	15	vision	vision	NOUN
fuelectenerg-13681	393	16	of	of	ADP
fuelectenerg-13681	393	17	6	6	NUM
fuelectenerg-13681	393	18	g	g	NOUN
fuelectenerg-13681	393	19	wireless	wireless	NOUN
fuelectenerg-13681	393	20	systems	system	NOUN
fuelectenerg-13681	393	21	:	:	PUNCT
fuelectenerg-13681	393	22	applications	application	NOUN
fuelectenerg-13681	393	23	,	,	PUNCT
fuelectenerg-13681	393	24	trends	trend	NOUN
fuelectenerg-13681	393	25	,	,	PUNCT
fuelectenerg-13681	393	26	technologies	technology	NOUN
fuelectenerg-13681	393	27	,	,	PUNCT
fuelectenerg-13681	393	28	and	and	CCONJ
fuelectenerg-13681	393	29	open	open	ADJ
fuelectenerg-13681	393	30	research	research	NOUN
fuelectenerg-13681	393	31	problems	problem	NOUN
fuelectenerg-13681	393	32	,	,	PUNCT
fuelectenerg-13681	393	33	"	"	PUNCT
fuelectenerg-13681	393	34	ieee	ieee	NOUN
fuelectenerg-13681	393	35	network	network	NOUN
fuelectenerg-13681	393	36	,	,	PUNCT
fuelectenerg-13681	393	37	vol	vol	NOUN
fuelectenerg-13681	393	38	.	.	PROPN
fuelectenerg-13681	393	39	34	34	NUM
fuelectenerg-13681	393	40	,	,	PUNCT
fuelectenerg-13681	393	41	no	no	INTJ
fuelectenerg-13681	393	42	.	.	NOUN
fuelectenerg-13681	393	43	3	3	NUM
fuelectenerg-13681	393	44	,	,	PUNCT
fuelectenerg-13681	393	45	pp	pp	ADJ
fuelectenerg-13681	393	46	.	.	PUNCT
fuelectenerg-13681	394	1	134–142	134–142	NUM
fuelectenerg-13681	394	2	,	,	PUNCT
fuelectenerg-13681	394	3	2020	2020	NUM
fuelectenerg-13681	394	4	.	.	PUNCT
fuelectenerg-13681	395	1	[	[	X
fuelectenerg-13681	395	2	4	4	X
fuelectenerg-13681	395	3	]	]	PUNCT
fuelectenerg-13681	395	4	l.	l.	PROPN
fuelectenerg-13681	395	5	liu	liu	PROPN
fuelectenerg-13681	395	6	,	,	PUNCT
fuelectenerg-13681	395	7	et	et	PROPN
fuelectenerg-13681	395	8	al	al	PROPN
fuelectenerg-13681	395	9	.	.	PROPN
fuelectenerg-13681	395	10	,	,	PUNCT
fuelectenerg-13681	395	11	"	"	PUNCT
fuelectenerg-13681	395	12	what	what	PRON
fuelectenerg-13681	395	13	is	be	AUX
fuelectenerg-13681	395	14	the	the	DET
fuelectenerg-13681	395	15	6	6	NUM
fuelectenerg-13681	395	16	g	g	NOUN
fuelectenerg-13681	395	17	vision	vision	NOUN
fuelectenerg-13681	395	18	?	?	PUNCT
fuelectenerg-13681	396	1	key	key	ADJ
fuelectenerg-13681	396	2	services	service	NOUN
fuelectenerg-13681	396	3	,	,	PUNCT
fuelectenerg-13681	396	4	use	use	VERB
fuelectenerg-13681	396	5	cases	case	NOUN
fuelectenerg-13681	396	6	,	,	PUNCT
fuelectenerg-13681	396	7	requirements	requirement	NOUN
fuelectenerg-13681	396	8	,	,	PUNCT
fuelectenerg-13681	396	9	and	and	CCONJ
fuelectenerg-13681	396	10	technologies	technology	NOUN
fuelectenerg-13681	396	11	,	,	PUNCT
fuelectenerg-13681	396	12	"	"	PUNCT
fuelectenerg-13681	396	13	philos	philo	NOUN
fuelectenerg-13681	396	14	.	.	PUNCT
fuelectenerg-13681	397	1	trans	trans	PROPN
fuelectenerg-13681	397	2	.	.	PUNCT
fuelectenerg-13681	398	1	royal	royal	ADJ
fuelectenerg-13681	398	2	soc	soc	PROPN
fuelectenerg-13681	398	3	.	.	PUNCT
fuelectenerg-13681	399	1	a	a	DET
fuelectenerg-13681	399	2	,	,	PUNCT
fuelectenerg-13681	399	3	vol	vol	NOUN
fuelectenerg-13681	399	4	.	.	PROPN
fuelectenerg-13681	399	5	380	380	NUM
fuelectenerg-13681	399	6	,	,	PUNCT
fuelectenerg-13681	399	7	no	no	INTJ
fuelectenerg-13681	399	8	.	.	NOUN
fuelectenerg-13681	399	9	2230	2230	NUM
fuelectenerg-13681	399	10	,	,	PUNCT
fuelectenerg-13681	399	11	2022	2022	NUM
fuelectenerg-13681	399	12	.	.	PUNCT
fuelectenerg-13681	400	1	[	[	X
fuelectenerg-13681	400	2	5	5	X
fuelectenerg-13681	400	3	]	]	X
fuelectenerg-13681	400	4	y.	y.	PROPN
fuelectenerg-13681	400	5	niu	niu	PROPN
fuelectenerg-13681	400	6	,	,	PUNCT
fuelectenerg-13681	400	7	y.	y.	PROPN
fuelectenerg-13681	400	8	li	li	PROPN
fuelectenerg-13681	400	9	,	,	PUNCT
fuelectenerg-13681	400	10	d.	d.	PROPN
fuelectenerg-13681	400	11	jin	jin	PROPN
fuelectenerg-13681	400	12	,	,	PUNCT
fuelectenerg-13681	400	13	l.	l.	PROPN
fuelectenerg-13681	400	14	su	su	PROPN
fuelectenerg-13681	400	15	and	and	CCONJ
fuelectenerg-13681	400	16	a.	a.	PROPN
fuelectenerg-13681	400	17	v.	v.	PROPN
fuelectenerg-13681	400	18	vasilakos	vasilakos	PROPN
fuelectenerg-13681	400	19	,	,	PUNCT
fuelectenerg-13681	400	20	"	"	PUNCT
fuelectenerg-13681	400	21	a	a	DET
fuelectenerg-13681	400	22	survey	survey	NOUN
fuelectenerg-13681	400	23	of	of	ADP
fuelectenerg-13681	400	24	millimeter	millimeter	NOUN
fuelectenerg-13681	400	25	wave	wave	NOUN
fuelectenerg-13681	400	26	communications	communication	NOUN
fuelectenerg-13681	400	27	(	(	PUNCT
fuelectenerg-13681	400	28	mmwave	mmwave	NOUN
fuelectenerg-13681	400	29	)	)	PUNCT
fuelectenerg-13681	400	30	for	for	ADP
fuelectenerg-13681	400	31	5	5	NUM
fuelectenerg-13681	400	32	g	g	NOUN
fuelectenerg-13681	400	33	:	:	PUNCT
fuelectenerg-13681	400	34	opportunities	opportunity	NOUN
fuelectenerg-13681	400	35	and	and	CCONJ
fuelectenerg-13681	400	36	challenges	challenge	NOUN
fuelectenerg-13681	400	37	,	,	PUNCT
fuelectenerg-13681	400	38	"	"	PUNCT
fuelectenerg-13681	400	39	wirel	wirel	NOUN
fuelectenerg-13681	400	40	.	.	PUNCT
fuelectenerg-13681	401	1	netw	netw	NOUN
fuelectenerg-13681	401	2	.	.	PUNCT
fuelectenerg-13681	402	1	,	,	PUNCT
fuelectenerg-13681	402	2	vol	vol	NOUN
fuelectenerg-13681	402	3	.	.	PROPN
fuelectenerg-13681	403	1	21	21	NUM
fuelectenerg-13681	403	2	,	,	PUNCT
fuelectenerg-13681	403	3	no	no	INTJ
fuelectenerg-13681	403	4	.	.	NOUN
fuelectenerg-13681	403	5	8	8	NUM
fuelectenerg-13681	403	6	,	,	PUNCT
fuelectenerg-13681	403	7	pp	pp	ADJ
fuelectenerg-13681	403	8	.	.	PUNCT
fuelectenerg-13681	404	1	2657–2676	2657–2676	NUM
fuelectenerg-13681	404	2	,	,	PUNCT
fuelectenerg-13681	404	3	2015	2015	NUM
fuelectenerg-13681	404	4	.	.	PUNCT
fuelectenerg-13681	405	1	[	[	X
fuelectenerg-13681	405	2	6	6	NUM
fuelectenerg-13681	405	3	]	]	PUNCT
fuelectenerg-13681	405	4	z.	z.	PROPN
fuelectenerg-13681	405	5	lin	lin	PROPN
fuelectenerg-13681	405	6	,	,	PUNCT
fuelectenerg-13681	405	7	t.	t.	PROPN
fuelectenerg-13681	405	8	yang	yang	PROPN
fuelectenerg-13681	405	9	,	,	PUNCT
fuelectenerg-13681	405	10	l.	l.	PROPN
fuelectenerg-13681	405	11	qin	qin	PROPN
fuelectenerg-13681	405	12	,	,	PUNCT
fuelectenerg-13681	405	13	w.	w.	PROPN
fuelectenerg-13681	405	14	guo	guo	PROPN
fuelectenerg-13681	405	15	and	and	CCONJ
fuelectenerg-13681	405	16	r.	r.	PROPN
fuelectenerg-13681	405	17	he	he	PRON
fuelectenerg-13681	405	18	,	,	PUNCT
fuelectenerg-13681	405	19	"	"	PUNCT
fuelectenerg-13681	405	20	beamforming	beamforme	VERB
fuelectenerg-13681	405	21	designs	design	NOUN
fuelectenerg-13681	405	22	and	and	CCONJ
fuelectenerg-13681	405	23	performance	performance	NOUN
fuelectenerg-13681	405	24	evaluations	evaluation	NOUN
fuelectenerg-13681	405	25	for	for	ADP
fuelectenerg-13681	405	26	intelligent	intelligent	ADJ
fuelectenerg-13681	405	27	reflecting	reflect	VERB
fuelectenerg-13681	405	28	surface	surface	NOUN
fuelectenerg-13681	405	29	-	-	PUNCT
fuelectenerg-13681	405	30	assisted	assist	VERB
fuelectenerg-13681	405	31	wireless	wireless	ADJ
fuelectenerg-13681	405	32	communications	communication	NOUN
fuelectenerg-13681	405	33	,	,	PUNCT
fuelectenerg-13681	405	34	"	"	PUNCT
fuelectenerg-13681	405	35	ieee	ieee	PROPN
fuelectenerg-13681	405	36	open	open	PROPN
fuelectenerg-13681	405	37	j.	j.	PROPN
fuelectenerg-13681	405	38	commun	commun	PROPN
fuelectenerg-13681	405	39	.	.	PUNCT
fuelectenerg-13681	406	1	soc	soc	PROPN
fuelectenerg-13681	406	2	.	.	PUNCT
fuelectenerg-13681	406	3	,	,	PUNCT
fuelectenerg-13681	406	4	vol	vol	NOUN
fuelectenerg-13681	406	5	.	.	PROPN
fuelectenerg-13681	407	1	3	3	NUM
fuelectenerg-13681	407	2	,	,	PUNCT
fuelectenerg-13681	407	3	pp	pp	ADJ
fuelectenerg-13681	407	4	.	.	PUNCT
fuelectenerg-13681	408	1	45–64	45–64	NUM
fuelectenerg-13681	408	2	,	,	PUNCT
fuelectenerg-13681	408	3	2022	2022	NUM
fuelectenerg-13681	408	4	.	.	PUNCT
fuelectenerg-13681	409	1	[	[	X
fuelectenerg-13681	409	2	7	7	X
fuelectenerg-13681	409	3	]	]	PUNCT
fuelectenerg-13681	409	4	s.	s.	PROPN
fuelectenerg-13681	409	5	a.	a.	PROPN
fuelectenerg-13681	409	6	busari	busari	PROPN
fuelectenerg-13681	409	7	,	,	PUNCT
fuelectenerg-13681	409	8	k.	k.	PROPN
fuelectenerg-13681	409	9	m.	m.	PROPN
fuelectenerg-13681	409	10	s.	s.	PROPN
fuelectenerg-13681	409	11	huq	huq	PROPN
fuelectenerg-13681	409	12	,	,	PUNCT
fuelectenerg-13681	409	13	s.	s.	PROPN
fuelectenerg-13681	409	14	mumtaz	mumtaz	PROPN
fuelectenerg-13681	409	15	,	,	PUNCT
fuelectenerg-13681	409	16	l.	l.	PROPN
fuelectenerg-13681	409	17	dai	dai	PROPN
fuelectenerg-13681	409	18	and	and	CCONJ
fuelectenerg-13681	409	19	j.	j.	PROPN
fuelectenerg-13681	409	20	rodriguez	rodriguez	PROPN
fuelectenerg-13681	409	21	,	,	PUNCT
fuelectenerg-13681	409	22	"	"	PUNCT
fuelectenerg-13681	409	23	millimeter	millimeter	NOUN
fuelectenerg-13681	409	24	-	-	PUNCT
fuelectenerg-13681	409	25	wave	wave	NOUN
fuelectenerg-13681	409	26	massive	massive	ADJ
fuelectenerg-13681	409	27	mimo	mimo	PROPN
fuelectenerg-13681	409	28	communication	communication	NOUN
fuelectenerg-13681	409	29	for	for	ADP
fuelectenerg-13681	409	30	future	future	ADJ
fuelectenerg-13681	409	31	wireless	wireless	ADJ
fuelectenerg-13681	409	32	systems	system	NOUN
fuelectenerg-13681	409	33	:	:	PUNCT
fuelectenerg-13681	409	34	a	a	DET
fuelectenerg-13681	409	35	survey	survey	NOUN
fuelectenerg-13681	409	36	,	,	PUNCT
fuelectenerg-13681	409	37	"	"	PUNCT
fuelectenerg-13681	409	38	ieee	ieee	X
fuelectenerg-13681	409	39	commun	commun	PROPN
fuelectenerg-13681	409	40	.	.	PUNCT
fuelectenerg-13681	410	1	surv	surv	PROPN
fuelectenerg-13681	410	2	.	.	PUNCT
fuelectenerg-13681	411	1	tutor	tutor	NOUN
fuelectenerg-13681	411	2	.	.	PUNCT
fuelectenerg-13681	412	1	,	,	PUNCT
fuelectenerg-13681	412	2	vol	vol	NOUN
fuelectenerg-13681	412	3	.	.	PROPN
fuelectenerg-13681	412	4	20	20	NUM
fuelectenerg-13681	412	5	,	,	PUNCT
fuelectenerg-13681	412	6	no	no	INTJ
fuelectenerg-13681	412	7	.	.	NOUN
fuelectenerg-13681	412	8	2	2	NUM
fuelectenerg-13681	412	9	,	,	PUNCT
fuelectenerg-13681	412	10	pp	pp	ADJ
fuelectenerg-13681	412	11	.	.	PUNCT
fuelectenerg-13681	413	1	836–869	836–869	NUM
fuelectenerg-13681	413	2	,	,	PUNCT
fuelectenerg-13681	413	3	2018	2018	NUM
fuelectenerg-13681	413	4	.	.	PUNCT
fuelectenerg-13681	414	1	[	[	X
fuelectenerg-13681	414	2	8	8	NUM
fuelectenerg-13681	414	3	]	]	X
fuelectenerg-13681	414	4	f.	f.	PROPN
fuelectenerg-13681	414	5	meng	meng	PROPN
fuelectenerg-13681	414	6	,	,	PUNCT
fuelectenerg-13681	414	7	k.	k.	PROPN
fuelectenerg-13681	414	8	cheng	cheng	PROPN
fuelectenerg-13681	414	9	,	,	PUNCT
fuelectenerg-13681	414	10	y.	y.	PROPN
fuelectenerg-13681	414	11	yang	yang	PROPN
fuelectenerg-13681	414	12	and	and	CCONJ
fuelectenerg-13681	414	13	z.	z.	PROPN
fuelectenerg-13681	414	14	wei	wei	PROPN
fuelectenerg-13681	414	15	,	,	PUNCT
fuelectenerg-13681	414	16	"	"	PUNCT
fuelectenerg-13681	414	17	deep	deep	ADJ
fuelectenerg-13681	414	18	reinforcement	reinforcement	NOUN
fuelectenerg-13681	414	19	learning	learning	NOUN
fuelectenerg-13681	414	20	-	-	PUNCT
fuelectenerg-13681	414	21	based	base	VERB
fuelectenerg-13681	414	22	power	power	NOUN
fuelectenerg-13681	414	23	control	control	NOUN
fuelectenerg-13681	414	24	for	for	ADP
fuelectenerg-13681	414	25	cellfree	cellfree	ADJ
fuelectenerg-13681	414	26	massive	massive	PROPN
fuelectenerg-13681	414	27	mimo	mimo	PROPN
fuelectenerg-13681	414	28	networks	network	NOUN
fuelectenerg-13681	414	29	,	,	PUNCT
fuelectenerg-13681	414	30	"	"	PUNCT
fuelectenerg-13681	414	31	ieee	ieee	NOUN
fuelectenerg-13681	414	32	trans	tran	NOUN
fuelectenerg-13681	414	33	.	.	PUNCT
fuelectenerg-13681	414	34	wirel	wirel	PROPN
fuelectenerg-13681	414	35	.	.	PUNCT
fuelectenerg-13681	415	1	commun	commun	PROPN
fuelectenerg-13681	415	2	.	.	PROPN
fuelectenerg-13681	415	3	,	,	PUNCT
fuelectenerg-13681	415	4	vol	vol	NOUN
fuelectenerg-13681	415	5	.	.	PROPN
fuelectenerg-13681	415	6	21	21	NUM
fuelectenerg-13681	415	7	,	,	PUNCT
fuelectenerg-13681	415	8	no	no	INTJ
fuelectenerg-13681	415	9	.	.	NOUN
fuelectenerg-13681	415	10	5	5	NUM
fuelectenerg-13681	415	11	,	,	PUNCT
fuelectenerg-13681	415	12	pp	pp	ADJ
fuelectenerg-13681	415	13	.	.	PUNCT
fuelectenerg-13681	416	1	3188–3202	3188–3202	NUM
fuelectenerg-13681	416	2	,	,	PUNCT
fuelectenerg-13681	416	3	2022	2022	NUM
fuelectenerg-13681	416	4	.	.	PUNCT
fuelectenerg-13681	417	1	[	[	X
fuelectenerg-13681	417	2	9	9	NUM
fuelectenerg-13681	417	3	]	]	PUNCT
fuelectenerg-13681	417	4	k.	k.	PROPN
fuelectenerg-13681	417	5	b.	b.	PROPN
fuelectenerg-13681	417	6	letaief	letaief	PROPN
fuelectenerg-13681	417	7	,	,	PUNCT
fuelectenerg-13681	417	8	w.	w.	PROPN
fuelectenerg-13681	417	9	chen	chen	PROPN
fuelectenerg-13681	417	10	,	,	PUNCT
fuelectenerg-13681	417	11	y.	y.	PROPN
fuelectenerg-13681	417	12	shi	shi	PROPN
fuelectenerg-13681	417	13	,	,	PUNCT
fuelectenerg-13681	417	14	j.	j.	PROPN
fuelectenerg-13681	417	15	zhang	zhang	PROPN
fuelectenerg-13681	417	16	and	and	CCONJ
fuelectenerg-13681	417	17	y.-j	y.-j	PROPN
fuelectenerg-13681	417	18	.	.	PUNCT
fuelectenerg-13681	417	19	a.	a.	PROPN
fuelectenerg-13681	417	20	zhang	zhang	PROPN
fuelectenerg-13681	417	21	,	,	PUNCT
fuelectenerg-13681	417	22	"	"	PUNCT
fuelectenerg-13681	417	23	the	the	DET
fuelectenerg-13681	417	24	roadmap	roadmap	NOUN
fuelectenerg-13681	417	25	to	to	ADP
fuelectenerg-13681	417	26	6	6	NUM
fuelectenerg-13681	417	27	g	g	NOUN
fuelectenerg-13681	417	28	:	:	PUNCT
fuelectenerg-13681	417	29	ai	ai	VERB
fuelectenerg-13681	417	30	empowered	empower	VERB
fuelectenerg-13681	417	31	wireless	wireless	ADJ
fuelectenerg-13681	417	32	networks	network	NOUN
fuelectenerg-13681	417	33	,	,	PUNCT
fuelectenerg-13681	417	34	"	"	PUNCT
fuelectenerg-13681	417	35	ieee	ieee	X
fuelectenerg-13681	417	36	commun	commun	PROPN
fuelectenerg-13681	417	37	.	.	PUNCT
fuelectenerg-13681	418	1	mag	mag	PROPN
fuelectenerg-13681	418	2	.	.	PROPN
fuelectenerg-13681	418	3	,	,	PUNCT
fuelectenerg-13681	418	4	vol	vol	NOUN
fuelectenerg-13681	418	5	.	.	PROPN
fuelectenerg-13681	419	1	57	57	NUM
fuelectenerg-13681	419	2	,	,	PUNCT
fuelectenerg-13681	419	3	no	no	INTJ
fuelectenerg-13681	419	4	.	.	NOUN
fuelectenerg-13681	419	5	8	8	NUM
fuelectenerg-13681	419	6	,	,	PUNCT
fuelectenerg-13681	419	7	pp	pp	ADJ
fuelectenerg-13681	419	8	.	.	PUNCT
fuelectenerg-13681	420	1	84–90	84–90	NUM
fuelectenerg-13681	420	2	,	,	PUNCT
fuelectenerg-13681	420	3	2019	2019	NUM
fuelectenerg-13681	420	4	.	.	PUNCT
fuelectenerg-13681	421	1	[	[	X
fuelectenerg-13681	421	2	10	10	NUM
fuelectenerg-13681	421	3	]	]	PUNCT
fuelectenerg-13681	421	4	a.	a.	NOUN
fuelectenerg-13681	421	5	zappone	zappone	NOUN
fuelectenerg-13681	421	6	,	,	PUNCT
fuelectenerg-13681	421	7	m.	m.	NOUN
fuelectenerg-13681	421	8	di	di	NOUN
fuelectenerg-13681	421	9	renzo	renzo	PROPN
fuelectenerg-13681	421	10	and	and	CCONJ
fuelectenerg-13681	421	11	m.	m.	NOUN
fuelectenerg-13681	421	12	debbah	debbah	PROPN
fuelectenerg-13681	421	13	,	,	PUNCT
fuelectenerg-13681	421	14	"	"	PUNCT
fuelectenerg-13681	421	15	wireless	wireless	ADJ
fuelectenerg-13681	421	16	networks	network	NOUN
fuelectenerg-13681	421	17	design	design	NOUN
fuelectenerg-13681	421	18	in	in	ADP
fuelectenerg-13681	421	19	the	the	DET
fuelectenerg-13681	421	20	era	era	NOUN
fuelectenerg-13681	421	21	of	of	ADP
fuelectenerg-13681	421	22	deep	deep	ADJ
fuelectenerg-13681	421	23	learning	learning	NOUN
fuelectenerg-13681	421	24	:	:	PUNCT
fuelectenerg-13681	421	25	model	model	NOUN
fuelectenerg-13681	421	26	-	-	PUNCT
fuelectenerg-13681	421	27	based	base	VERB
fuelectenerg-13681	421	28	,	,	PUNCT
fuelectenerg-13681	421	29	ai	ai	NOUN
fuelectenerg-13681	421	30	-	-	PUNCT
fuelectenerg-13681	421	31	based	base	VERB
fuelectenerg-13681	421	32	,	,	PUNCT
fuelectenerg-13681	421	33	or	or	CCONJ
fuelectenerg-13681	421	34	both	both	PRON
fuelectenerg-13681	421	35	?	?	PUNCT
fuelectenerg-13681	421	36	,	,	PUNCT
fuelectenerg-13681	421	37	"	"	PUNCT
fuelectenerg-13681	421	38	ieee	ieee	NOUN
fuelectenerg-13681	421	39	trans	tran	NOUN
fuelectenerg-13681	421	40	.	.	PUNCT
fuelectenerg-13681	422	1	commun	commun	PROPN
fuelectenerg-13681	422	2	.	.	PROPN
fuelectenerg-13681	422	3	,	,	PUNCT
fuelectenerg-13681	422	4	vol	vol	NOUN
fuelectenerg-13681	422	5	.	.	PROPN
fuelectenerg-13681	422	6	67	67	NUM
fuelectenerg-13681	422	7	,	,	PUNCT
fuelectenerg-13681	422	8	no	no	INTJ
fuelectenerg-13681	422	9	.	.	NOUN
fuelectenerg-13681	422	10	10	10	NUM
fuelectenerg-13681	422	11	,	,	PUNCT
fuelectenerg-13681	422	12	pp	pp	ADJ
fuelectenerg-13681	422	13	.	.	PUNCT
fuelectenerg-13681	423	1	7178–7195	7178–7195	NUM
fuelectenerg-13681	423	2	,	,	PUNCT
fuelectenerg-13681	423	3	2019	2019	NUM
fuelectenerg-13681	423	4	.	.	PUNCT
fuelectenerg-13681	424	1	[	[	X
fuelectenerg-13681	424	2	11	11	NUM
fuelectenerg-13681	424	3	]	]	PUNCT
fuelectenerg-13681	424	4	k.	k.	PROPN
fuelectenerg-13681	424	5	s.	s.	PROPN
fuelectenerg-13681	424	6	khan	khan	PROPN
fuelectenerg-13681	424	7	and	and	CCONJ
fuelectenerg-13681	424	8	j.	j.	PROPN
fuelectenerg-13681	424	9	usman	usman	PROPN
fuelectenerg-13681	424	10	,	,	PUNCT
fuelectenerg-13681	424	11	"	"	PUNCT
fuelectenerg-13681	424	12	backtracking	backtrack	VERB
fuelectenerg-13681	424	13	search	search	NOUN
fuelectenerg-13681	424	14	optimization	optimization	NOUN
fuelectenerg-13681	424	15	algorithm	algorithm	NOUN
fuelectenerg-13681	424	16	for	for	ADP
fuelectenerg-13681	424	17	synthesis	synthesis	NOUN
fuelectenerg-13681	424	18	of	of	ADP
fuelectenerg-13681	424	19	linear	linear	ADJ
fuelectenerg-13681	424	20	antenna	antenna	NOUN
fuelectenerg-13681	424	21	arrays	array	VERB
fuelectenerg-13681	424	22	with	with	ADP
fuelectenerg-13681	424	23	multi	multi	ADJ
fuelectenerg-13681	424	24	-	-	ADJ
fuelectenerg-13681	424	25	objective	objective	ADJ
fuelectenerg-13681	424	26	functions	function	NOUN
fuelectenerg-13681	424	27	,	,	PUNCT
fuelectenerg-13681	424	28	"	"	PUNCT
fuelectenerg-13681	424	29	ieee	ieee	NOUN
fuelectenerg-13681	424	30	trans	tran	NOUN
fuelectenerg-13681	424	31	.	.	PUNCT
fuelectenerg-13681	425	1	antennas	antennas	PROPN
fuelectenerg-13681	425	2	propag	propag	PROPN
fuelectenerg-13681	425	3	.	.	PUNCT
fuelectenerg-13681	425	4	,	,	PUNCT
fuelectenerg-13681	425	5	vol	vol	NOUN
fuelectenerg-13681	425	6	.	.	PROPN
fuelectenerg-13681	426	1	70	70	NUM
fuelectenerg-13681	426	2	,	,	PUNCT
fuelectenerg-13681	426	3	no	no	INTJ
fuelectenerg-13681	426	4	.	.	NOUN
fuelectenerg-13681	426	5	2	2	NUM
fuelectenerg-13681	426	6	,	,	PUNCT
fuelectenerg-13681	426	7	pp	pp	ADJ
fuelectenerg-13681	426	8	.	.	PUNCT
fuelectenerg-13681	427	1	1110–1119	1110–1119	NUM
fuelectenerg-13681	427	2	,	,	PUNCT
fuelectenerg-13681	427	3	2022	2022	NUM
fuelectenerg-13681	427	4	.	.	PUNCT
fuelectenerg-13681	428	1	[	[	X
fuelectenerg-13681	428	2	12	12	NUM
fuelectenerg-13681	428	3	]	]	PUNCT
fuelectenerg-13681	428	4	j.	j.	PROPN
fuelectenerg-13681	428	5	zhang	zhang	PROPN
fuelectenerg-13681	428	6	,	,	PUNCT
fuelectenerg-13681	428	7	c.	c.	PROPN
fuelectenerg-13681	428	8	chen	chen	PROPN
fuelectenerg-13681	428	9	,	,	PUNCT
fuelectenerg-13681	428	10	x.	x.	PROPN
fuelectenerg-13681	428	11	pei	pei	PROPN
fuelectenerg-13681	428	12	and	and	CCONJ
fuelectenerg-13681	428	13	d.	d.	PROPN
fuelectenerg-13681	428	14	xue	xue	PROPN
fuelectenerg-13681	428	15	,	,	PUNCT
fuelectenerg-13681	428	16	"	"	PUNCT
fuelectenerg-13681	428	17	multi	multi	ADJ
fuelectenerg-13681	428	18	-	-	NOUN
fuelectenerg-13681	428	19	user	user	ADJ
fuelectenerg-13681	428	20	power	power	NOUN
fuelectenerg-13681	428	21	control	control	NOUN
fuelectenerg-13681	428	22	algorithm	algorithm	NOUN
fuelectenerg-13681	428	23	based	base	VERB
fuelectenerg-13681	428	24	on	on	ADP
fuelectenerg-13681	428	25	deep	deep	ADJ
fuelectenerg-13681	428	26	reinforcement	reinforcement	NOUN
fuelectenerg-13681	428	27	learning	learning	NOUN
fuelectenerg-13681	428	28	in	in	ADP
fuelectenerg-13681	428	29	fdd	fdd	PROPN
fuelectenerg-13681	428	30	massive	massive	PROPN
fuelectenerg-13681	428	31	mimo	mimo	PROPN
fuelectenerg-13681	428	32	systems	systems	PROPN
fuelectenerg-13681	428	33	,	,	PUNCT
fuelectenerg-13681	428	34	"	"	PUNCT
fuelectenerg-13681	428	35	ieee	ieee	X
fuelectenerg-13681	428	36	commun	commun	PROPN
fuelectenerg-13681	428	37	.	.	PUNCT
fuelectenerg-13681	429	1	lett	lett	PROPN
fuelectenerg-13681	429	2	.	.	PROPN
fuelectenerg-13681	429	3	,	,	PUNCT
fuelectenerg-13681	429	4	vol	vol	NOUN
fuelectenerg-13681	429	5	.	.	PROPN
fuelectenerg-13681	430	1	24	24	NUM
fuelectenerg-13681	430	2	,	,	PUNCT
fuelectenerg-13681	430	3	no	no	INTJ
fuelectenerg-13681	430	4	.	.	NOUN
fuelectenerg-13681	430	5	8	8	NUM
fuelectenerg-13681	430	6	,	,	PUNCT
fuelectenerg-13681	430	7	pp	pp	ADJ
fuelectenerg-13681	430	8	.	.	PUNCT
fuelectenerg-13681	431	1	1753–1757	1753–1757	NUM
fuelectenerg-13681	431	2	,	,	PUNCT
fuelectenerg-13681	431	3	2020	2020	NUM
fuelectenerg-13681	431	4	.	.	PUNCT
fuelectenerg-13681	432	1	[	[	X
fuelectenerg-13681	432	2	13	13	NUM
fuelectenerg-13681	432	3	]	]	PUNCT
fuelectenerg-13681	432	4	a.	a.	NOUN
fuelectenerg-13681	432	5	m.	m.	NOUN
fuelectenerg-13681	432	6	elbir	elbir	NOUN
fuelectenerg-13681	432	7	and	and	CCONJ
fuelectenerg-13681	432	8	k.	k.	PROPN
fuelectenerg-13681	432	9	v.	v.	PROPN
fuelectenerg-13681	432	10	mishra	mishra	PROPN
fuelectenerg-13681	432	11	,	,	PUNCT
fuelectenerg-13681	432	12	"	"	PUNCT
fuelectenerg-13681	432	13	a	a	DET
fuelectenerg-13681	432	14	survey	survey	NOUN
fuelectenerg-13681	432	15	of	of	ADP
fuelectenerg-13681	432	16	deep	deep	ADJ
fuelectenerg-13681	432	17	learning	learning	NOUN
fuelectenerg-13681	432	18	architectures	architecture	NOUN
fuelectenerg-13681	432	19	for	for	ADP
fuelectenerg-13681	432	20	intelligent	intelligent	ADJ
fuelectenerg-13681	432	21	reflecting	reflect	VERB
fuelectenerg-13681	432	22	surfaces	surface	NOUN
fuelectenerg-13681	432	23	,	,	PUNCT
fuelectenerg-13681	432	24	"	"	PUNCT
fuelectenerg-13681	432	25	ieee	ieee	X
fuelectenerg-13681	432	26	commun	commun	PROPN
fuelectenerg-13681	432	27	.	.	PUNCT
fuelectenerg-13681	433	1	surv	surv	PROPN
fuelectenerg-13681	433	2	.	.	PUNCT
fuelectenerg-13681	434	1	tutor	tutor	NOUN
fuelectenerg-13681	434	2	.	.	PUNCT
fuelectenerg-13681	435	1	,	,	PUNCT
fuelectenerg-13681	435	2	vol	vol	NOUN
fuelectenerg-13681	435	3	.	.	PROPN
fuelectenerg-13681	436	1	24	24	NUM
fuelectenerg-13681	436	2	,	,	PUNCT
fuelectenerg-13681	436	3	no	no	INTJ
fuelectenerg-13681	436	4	.	.	NOUN
fuelectenerg-13681	436	5	2	2	NUM
fuelectenerg-13681	436	6	,	,	PUNCT
fuelectenerg-13681	436	7	pp	pp	ADJ
fuelectenerg-13681	436	8	.	.	PUNCT
fuelectenerg-13681	437	1	1050–1091	1050–1091	NUM
fuelectenerg-13681	437	2	,	,	PUNCT
fuelectenerg-13681	437	3	2022	2022	NUM
fuelectenerg-13681	437	4	.	.	PUNCT
fuelectenerg-13681	438	1	[	[	X
fuelectenerg-13681	438	2	14	14	NUM
fuelectenerg-13681	438	3	]	]	X
fuelectenerg-13681	438	4	c.	c.	PROPN
fuelectenerg-13681	438	5	huang	huang	PROPN
fuelectenerg-13681	438	6	,	,	PUNCT
fuelectenerg-13681	438	7	r.	r.	PROPN
fuelectenerg-13681	438	8	mo	mo	PROPN
fuelectenerg-13681	438	9	,	,	PUNCT
fuelectenerg-13681	438	10	c.	c.	PROPN
fuelectenerg-13681	438	11	yuen	yuen	PROPN
fuelectenerg-13681	438	12	and	and	CCONJ
fuelectenerg-13681	438	13	a.	a.	NOUN
fuelectenerg-13681	438	14	nallanathan	nallanathan	PROPN
fuelectenerg-13681	438	15	,	,	PUNCT
fuelectenerg-13681	438	16	"	"	PUNCT
fuelectenerg-13681	438	17	deep	deep	ADJ
fuelectenerg-13681	438	18	reinforcement	reinforcement	NOUN
fuelectenerg-13681	438	19	learning	learning	NOUN
fuelectenerg-13681	438	20	-	-	PUNCT
fuelectenerg-13681	438	21	based	base	VERB
fuelectenerg-13681	438	22	beamforming	beamforming	NOUN
fuelectenerg-13681	438	23	and	and	CCONJ
fuelectenerg-13681	438	24	power	power	NOUN
fuelectenerg-13681	438	25	control	control	NOUN
fuelectenerg-13681	438	26	for	for	ADP
fuelectenerg-13681	438	27	irs	irs	PROPN
fuelectenerg-13681	438	28	-	-	PUNCT
fuelectenerg-13681	438	29	assisted	assist	VERB
fuelectenerg-13681	438	30	noma	noma	PROPN
fuelectenerg-13681	438	31	networks	network	NOUN
fuelectenerg-13681	438	32	,	,	PUNCT
fuelectenerg-13681	438	33	"	"	PUNCT
fuelectenerg-13681	438	34	ieee	ieee	NOUN
fuelectenerg-13681	438	35	trans	trans	PROPN
fuelectenerg-13681	438	36	.	.	PUNCT
fuelectenerg-13681	439	1	veh	veh	PROPN
fuelectenerg-13681	439	2	.	.	PROPN
fuelectenerg-13681	439	3	technol	technol	PROPN
fuelectenerg-13681	439	4	.	.	PROPN
fuelectenerg-13681	439	5	,	,	PUNCT
fuelectenerg-13681	439	6	vol	vol	NOUN
fuelectenerg-13681	439	7	.	.	PROPN
fuelectenerg-13681	440	1	71	71	NUM
fuelectenerg-13681	440	2	,	,	PUNCT
fuelectenerg-13681	440	3	no	no	INTJ
fuelectenerg-13681	440	4	.	.	NOUN
fuelectenerg-13681	440	5	8	8	NUM
fuelectenerg-13681	440	6	,	,	PUNCT
fuelectenerg-13681	440	7	pp	pp	ADJ
fuelectenerg-13681	440	8	.	.	PUNCT
fuelectenerg-13681	441	1	8795	8795	NUM
fuelectenerg-13681	441	2	–	–	PUNCT
fuelectenerg-13681	441	3	8808	8808	NUM
fuelectenerg-13681	441	4	,	,	PUNCT
fuelectenerg-13681	441	5	2022	2022	NUM
fuelectenerg-13681	441	6	.	.	PUNCT
fuelectenerg-13681	442	1	[	[	X
fuelectenerg-13681	442	2	15	15	NUM
fuelectenerg-13681	442	3	]	]	X
fuelectenerg-13681	442	4	j.	j.	PROPN
fuelectenerg-13681	442	5	guo	guo	PROPN
fuelectenerg-13681	442	6	and	and	CCONJ
fuelectenerg-13681	442	7	c.	c.	PROPN
fuelectenerg-13681	442	8	yang	yang	PROPN
fuelectenerg-13681	442	9	,	,	PUNCT
fuelectenerg-13681	442	10	"	"	PUNCT
fuelectenerg-13681	442	11	an	an	DET
fuelectenerg-13681	442	12	overview	overview	NOUN
fuelectenerg-13681	442	13	on	on	ADP
fuelectenerg-13681	442	14	deep	deep	ADJ
fuelectenerg-13681	442	15	learning	learning	NOUN
fuelectenerg-13681	442	16	-	-	PUNCT
fuelectenerg-13681	442	17	based	base	VERB
fuelectenerg-13681	442	18	wireless	wireless	ADJ
fuelectenerg-13681	442	19	signal	signal	NOUN
fuelectenerg-13681	442	20	recognition	recognition	NOUN
fuelectenerg-13681	442	21	and	and	CCONJ
fuelectenerg-13681	442	22	its	its	PRON
fuelectenerg-13681	442	23	recent	recent	ADJ
fuelectenerg-13681	442	24	advances	advance	NOUN
fuelectenerg-13681	442	25	,	,	PUNCT
fuelectenerg-13681	442	26	"	"	PUNCT
fuelectenerg-13681	442	27	int	int	NOUN
fuelectenerg-13681	442	28	.	.	PUNCT
fuelectenerg-13681	443	1	j.	j.	PROPN
fuelectenerg-13681	443	2	comput	comput	PROPN
fuelectenerg-13681	443	3	.	.	PUNCT
fuelectenerg-13681	444	1	intell	intell	PROPN
fuelectenerg-13681	444	2	.	.	PUNCT
fuelectenerg-13681	445	1	syst	syst	PROPN
fuelectenerg-13681	445	2	.	.	PUNCT
fuelectenerg-13681	445	3	,	,	PUNCT
fuelectenerg-13681	445	4	vol	vol	NOUN
fuelectenerg-13681	445	5	.	.	PROPN
fuelectenerg-13681	446	1	14	14	NUM
fuelectenerg-13681	446	2	,	,	PUNCT
fuelectenerg-13681	446	3	no	no	INTJ
fuelectenerg-13681	446	4	.	.	NOUN
fuelectenerg-13681	446	5	1	1	NUM
fuelectenerg-13681	446	6	,	,	PUNCT
fuelectenerg-13681	446	7	pp	pp	ADJ
fuelectenerg-13681	446	8	.	.	PUNCT
fuelectenerg-13681	447	1	788–803	788–803	NUM
fuelectenerg-13681	447	2	,	,	PUNCT
fuelectenerg-13681	447	3	2021	2021	NUM
fuelectenerg-13681	447	4	.	.	PUNCT
fuelectenerg-13681	448	1	[	[	X
fuelectenerg-13681	448	2	16	16	NUM
fuelectenerg-13681	448	3	]	]	X
fuelectenerg-13681	448	4	h.	h.	PROPN
fuelectenerg-13681	448	5	ye	ye	PROPN
fuelectenerg-13681	448	6	,	,	PUNCT
fuelectenerg-13681	448	7	g.	g.	PROPN
fuelectenerg-13681	448	8	y.	y.	PROPN
fuelectenerg-13681	448	9	li	li	PROPN
fuelectenerg-13681	448	10	and	and	CCONJ
fuelectenerg-13681	448	11	b.-h	b.-h	PROPN
fuelectenerg-13681	448	12	.	.	PUNCT
fuelectenerg-13681	449	1	f.	f.	PROPN
fuelectenerg-13681	449	2	juang	juang	PROPN
fuelectenerg-13681	449	3	,	,	PUNCT
fuelectenerg-13681	449	4	"	"	PUNCT
fuelectenerg-13681	449	5	power	power	NOUN
fuelectenerg-13681	449	6	of	of	ADP
fuelectenerg-13681	449	7	deep	deep	ADJ
fuelectenerg-13681	449	8	learning	learning	NOUN
fuelectenerg-13681	449	9	for	for	ADP
fuelectenerg-13681	449	10	channel	channel	NOUN
fuelectenerg-13681	449	11	estimation	estimation	NOUN
fuelectenerg-13681	449	12	and	and	CCONJ
fuelectenerg-13681	449	13	signal	signal	ADJ
fuelectenerg-13681	449	14	detection	detection	NOUN
fuelectenerg-13681	449	15	in	in	ADP
fuelectenerg-13681	449	16	ofdm	ofdm	PROPN
fuelectenerg-13681	449	17	systems	system	NOUN
fuelectenerg-13681	449	18	,	,	PUNCT
fuelectenerg-13681	449	19	"	"	PUNCT
fuelectenerg-13681	449	20	ieee	ieee	NOUN
fuelectenerg-13681	449	21	wirel	wirel	NOUN
fuelectenerg-13681	449	22	.	.	PUNCT
fuelectenerg-13681	450	1	commun	commun	PROPN
fuelectenerg-13681	450	2	.	.	PUNCT
fuelectenerg-13681	451	1	lett	lett	PROPN
fuelectenerg-13681	451	2	.	.	PROPN
fuelectenerg-13681	451	3	,	,	PUNCT
fuelectenerg-13681	451	4	vol	vol	NOUN
fuelectenerg-13681	451	5	.	.	PROPN
fuelectenerg-13681	451	6	7	7	NUM
fuelectenerg-13681	451	7	,	,	PUNCT
fuelectenerg-13681	451	8	no	no	INTJ
fuelectenerg-13681	451	9	.	.	NOUN
fuelectenerg-13681	451	10	1	1	NUM
fuelectenerg-13681	451	11	,	,	PUNCT
fuelectenerg-13681	451	12	pp	pp	ADJ
fuelectenerg-13681	451	13	.	.	PUNCT
fuelectenerg-13681	452	1	114–117	114–117	NUM
fuelectenerg-13681	452	2	,	,	PUNCT
fuelectenerg-13681	452	3	2018	2018	NUM
fuelectenerg-13681	452	4	.	.	PUNCT
fuelectenerg-13681	453	1	[	[	X
fuelectenerg-13681	453	2	17	17	NUM
fuelectenerg-13681	453	3	]	]	PUNCT
fuelectenerg-13681	453	4	l.	l.	PROPN
fuelectenerg-13681	453	5	liu	liu	PROPN
fuelectenerg-13681	453	6	,	,	PUNCT
fuelectenerg-13681	453	7	c.	c.	PROPN
fuelectenerg-13681	453	8	oestges	oestges	PROPN
fuelectenerg-13681	453	9	,	,	PUNCT
fuelectenerg-13681	453	10	j.	j.	PROPN
fuelectenerg-13681	453	11	poutanen	poutanen	PROPN
fuelectenerg-13681	453	12	,	,	PUNCT
fuelectenerg-13681	453	13	k.	k.	PROPN
fuelectenerg-13681	453	14	haneda	haneda	PROPN
fuelectenerg-13681	453	15	,	,	PUNCT
fuelectenerg-13681	453	16	p.	p.	PROPN
fuelectenerg-13681	453	17	vainikainen	vainikainen	PROPN
fuelectenerg-13681	453	18	,	,	PUNCT
fuelectenerg-13681	453	19	f.	f.	PROPN
fuelectenerg-13681	453	20	quitin	quitin	PROPN
fuelectenerg-13681	453	21	,	,	PUNCT
fuelectenerg-13681	453	22	f.	f.	PROPN
fuelectenerg-13681	453	23	tufvesson	tufvesson	PROPN
fuelectenerg-13681	453	24	and	and	CCONJ
fuelectenerg-13681	453	25	p.	p.	PROPN
fuelectenerg-13681	453	26	de	de	PROPN
fuelectenerg-13681	453	27	doncker	doncker	NOUN
fuelectenerg-13681	453	28	,	,	PUNCT
fuelectenerg-13681	453	29	"	"	PUNCT
fuelectenerg-13681	453	30	the	the	DET
fuelectenerg-13681	453	31	cost	cost	NOUN
fuelectenerg-13681	453	32	2100	2100	NUM
fuelectenerg-13681	453	33	mimo	mimo	PROPN
fuelectenerg-13681	453	34	channel	channel	PROPN
fuelectenerg-13681	453	35	model	model	NOUN
fuelectenerg-13681	453	36	,	,	PUNCT
fuelectenerg-13681	453	37	"	"	PUNCT
fuelectenerg-13681	453	38	ieee	ieee	NOUN
fuelectenerg-13681	453	39	wirel	wirel	NOUN
fuelectenerg-13681	453	40	.	.	PUNCT
fuelectenerg-13681	454	1	commun	commun	PROPN
fuelectenerg-13681	454	2	.	.	PROPN
fuelectenerg-13681	454	3	,	,	PUNCT
fuelectenerg-13681	454	4	vol	vol	NOUN
fuelectenerg-13681	454	5	.	.	PROPN
fuelectenerg-13681	454	6	19	19	NUM
fuelectenerg-13681	454	7	,	,	PUNCT
fuelectenerg-13681	454	8	no	no	INTJ
fuelectenerg-13681	454	9	.	.	NOUN
fuelectenerg-13681	454	10	6	6	NUM
fuelectenerg-13681	454	11	,	,	PUNCT
fuelectenerg-13681	454	12	pp	pp	ADJ
fuelectenerg-13681	454	13	.	.	PUNCT
fuelectenerg-13681	455	1	92–99	92–99	NUM
fuelectenerg-13681	455	2	,	,	PUNCT
fuelectenerg-13681	455	3	2012	2012	NUM
fuelectenerg-13681	455	4	.	.	PUNCT
fuelectenerg-13681	456	1	[	[	X
fuelectenerg-13681	456	2	18	18	NUM
fuelectenerg-13681	456	3	]	]	PUNCT
fuelectenerg-13681	456	4	z.	z.	PROPN
fuelectenerg-13681	456	5	liu	liu	PROPN
fuelectenerg-13681	456	6	,	,	PUNCT
fuelectenerg-13681	456	7	l.	l.	PROPN
fuelectenerg-13681	456	8	lei	lei	PROPN
fuelectenerg-13681	456	9	,	,	PUNCT
fuelectenerg-13681	456	10	n.	n.	PROPN
fuelectenerg-13681	456	11	zhang	zhang	PROPN
fuelectenerg-13681	456	12	,	,	PUNCT
fuelectenerg-13681	456	13	g.	g.	PROPN
fuelectenerg-13681	456	14	kang	kang	PROPN
fuelectenerg-13681	456	15	and	and	CCONJ
fuelectenerg-13681	456	16	s.	s.	PROPN
fuelectenerg-13681	456	17	chatzinotas	chatzinotas	PROPN
fuelectenerg-13681	456	18	,	,	PUNCT
fuelectenerg-13681	456	19	"	"	PUNCT
fuelectenerg-13681	456	20	joint	joint	ADJ
fuelectenerg-13681	456	21	beamforming	beamforming	NOUN
fuelectenerg-13681	456	22	and	and	CCONJ
fuelectenerg-13681	456	23	power	power	NOUN
fuelectenerg-13681	456	24	optimization	optimization	NOUN
fuelectenerg-13681	456	25	with	with	ADP
fuelectenerg-13681	456	26	iterative	iterative	ADJ
fuelectenerg-13681	456	27	user	user	NOUN
fuelectenerg-13681	456	28	clustering	cluster	VERB
fuelectenerg-13681	456	29	for	for	ADP
fuelectenerg-13681	456	30	miso	miso	NOUN
fuelectenerg-13681	456	31	-	-	PUNCT
fuelectenerg-13681	456	32	noma	noma	NOUN
fuelectenerg-13681	456	33	systems	system	NOUN
fuelectenerg-13681	456	34	,	,	PUNCT
fuelectenerg-13681	456	35	"	"	PUNCT
fuelectenerg-13681	456	36	ieee	ieee	NOUN
fuelectenerg-13681	456	37	access	access	NOUN
fuelectenerg-13681	456	38	,	,	PUNCT
fuelectenerg-13681	456	39	vol	vol	NOUN
fuelectenerg-13681	456	40	.	.	PROPN
fuelectenerg-13681	456	41	8	8	NUM
fuelectenerg-13681	456	42	,	,	PUNCT
fuelectenerg-13681	456	43	pp	pp	ADJ
fuelectenerg-13681	456	44	.	.	PUNCT
fuelectenerg-13681	457	1	24288–24301	24288–24301	NUM
fuelectenerg-13681	457	2	,	,	PUNCT
fuelectenerg-13681	457	3	2020	2020	NUM
fuelectenerg-13681	457	4	.	.	PUNCT
fuelectenerg-13681	458	1	[	[	X
fuelectenerg-13681	458	2	19	19	NUM
fuelectenerg-13681	458	3	]	]	PUNCT
fuelectenerg-13681	458	4	a.	a.	NOUN
fuelectenerg-13681	458	5	k.	k.	PROPN
fuelectenerg-13681	458	6	chaudhary	chaudhary	PROPN
fuelectenerg-13681	458	7	,	,	PUNCT
fuelectenerg-13681	458	8	k.-c	k.-c	PROPN
fuelectenerg-13681	458	9	.	.	PUNCT
fuelectenerg-13681	459	1	chen	chen	PROPN
fuelectenerg-13681	459	2	and	and	CCONJ
fuelectenerg-13681	459	3	h.	h.	PROPN
fuelectenerg-13681	459	4	v.	v.	PROPN
fuelectenerg-13681	459	5	poor	poor	ADJ
fuelectenerg-13681	459	6	,	,	PUNCT
fuelectenerg-13681	459	7	"	"	PUNCT
fuelectenerg-13681	459	8	deep	deep	ADJ
fuelectenerg-13681	459	9	learning	learning	NOUN
fuelectenerg-13681	459	10	based	base	VERB
fuelectenerg-13681	459	11	power	power	NOUN
fuelectenerg-13681	459	12	control	control	NOUN
fuelectenerg-13681	459	13	for	for	ADP
fuelectenerg-13681	459	14	cell	cell	NOUN
fuelectenerg-13681	459	15	-	-	PUNCT
fuelectenerg-13681	459	16	free	free	ADJ
fuelectenerg-13681	459	17	massive	massive	ADJ
fuelectenerg-13681	459	18	mimo	mimo	PROPN
fuelectenerg-13681	459	19	networks	network	NOUN
fuelectenerg-13681	459	20	,	,	PUNCT
fuelectenerg-13681	459	21	"	"	PUNCT
fuelectenerg-13681	459	22	ieee	ieee	NOUN
fuelectenerg-13681	459	23	trans	tran	NOUN
fuelectenerg-13681	459	24	.	.	PUNCT
fuelectenerg-13681	460	1	commun	commun	PROPN
fuelectenerg-13681	460	2	.	.	PROPN
fuelectenerg-13681	460	3	,	,	PUNCT
fuelectenerg-13681	460	4	vol	vol	NOUN
fuelectenerg-13681	460	5	.	.	PROPN
fuelectenerg-13681	461	1	70	70	NUM
fuelectenerg-13681	461	2	,	,	PUNCT
fuelectenerg-13681	461	3	no	no	INTJ
fuelectenerg-13681	461	4	.	.	NOUN
fuelectenerg-13681	461	5	4	4	NUM
fuelectenerg-13681	461	6	,	,	PUNCT
fuelectenerg-13681	461	7	pp	pp	ADJ
fuelectenerg-13681	461	8	.	.	PUNCT
fuelectenerg-13681	462	1	2615–2629	2615–2629	NUM
fuelectenerg-13681	462	2	,	,	PUNCT
fuelectenerg-13681	462	3	2022	2022	NUM
fuelectenerg-13681	462	4	.	.	PUNCT
fuelectenerg-13681	463	1	[	[	X
fuelectenerg-13681	463	2	20	20	NUM
fuelectenerg-13681	463	3	]	]	X
fuelectenerg-13681	463	4	w.	w.	PROPN
fuelectenerg-13681	463	5	wang	wang	PROPN
fuelectenerg-13681	463	6	,	,	PUNCT
fuelectenerg-13681	463	7	a.	a.	PROPN
fuelectenerg-13681	463	8	liu	liu	PROPN
fuelectenerg-13681	463	9	,	,	PUNCT
fuelectenerg-13681	463	10	q.	q.	PROPN
fuelectenerg-13681	463	11	zhang	zhang	PROPN
fuelectenerg-13681	463	12	,	,	PUNCT
fuelectenerg-13681	463	13	l.	l.	PROPN
fuelectenerg-13681	463	14	you	you	PRON
fuelectenerg-13681	463	15	,	,	PUNCT
fuelectenerg-13681	463	16	x.	x.	PROPN
fuelectenerg-13681	463	17	gao	gao	PROPN
fuelectenerg-13681	463	18	and	and	CCONJ
fuelectenerg-13681	463	19	g.	g.	PROPN
fuelectenerg-13681	463	20	zheng	zheng	PROPN
fuelectenerg-13681	463	21	,	,	PUNCT
fuelectenerg-13681	463	22	"	"	PUNCT
fuelectenerg-13681	463	23	fast	fast	ADJ
fuelectenerg-13681	463	24	and	and	CCONJ
fuelectenerg-13681	463	25	accurate	accurate	ADJ
fuelectenerg-13681	463	26	channel	channel	NOUN
fuelectenerg-13681	463	27	estimation	estimation	NOUN
fuelectenerg-13681	463	28	for	for	ADP
fuelectenerg-13681	463	29	mmwave	mmwave	PROPN
fuelectenerg-13681	463	30	massive	massive	PROPN
fuelectenerg-13681	463	31	mimo	mimo	PROPN
fuelectenerg-13681	463	32	systems	system	NOUN
fuelectenerg-13681	463	33	with	with	ADP
fuelectenerg-13681	463	34	one	one	NUM
fuelectenerg-13681	463	35	-	-	PUNCT
fuelectenerg-13681	463	36	bit	bit	NOUN
fuelectenerg-13681	463	37	adcs	adcs	ADJ
fuelectenerg-13681	463	38	,	,	PUNCT
fuelectenerg-13681	463	39	"	"	PUNCT
fuelectenerg-13681	463	40	ieee	ieee	NOUN
fuelectenerg-13681	463	41	trans	tran	NOUN
fuelectenerg-13681	463	42	.	.	PUNCT
fuelectenerg-13681	463	43	wirel	wirel	PROPN
fuelectenerg-13681	463	44	.	.	PUNCT
fuelectenerg-13681	464	1	commun	commun	PROPN
fuelectenerg-13681	464	2	.	.	PROPN
fuelectenerg-13681	464	3	,	,	PUNCT
fuelectenerg-13681	464	4	vol	vol	NOUN
fuelectenerg-13681	464	5	.	.	PROPN
fuelectenerg-13681	464	6	19	19	NUM
fuelectenerg-13681	464	7	,	,	PUNCT
fuelectenerg-13681	464	8	no	no	INTJ
fuelectenerg-13681	464	9	.	.	NOUN
fuelectenerg-13681	464	10	1	1	NUM
fuelectenerg-13681	464	11	,	,	PUNCT
fuelectenerg-13681	464	12	pp	pp	ADJ
fuelectenerg-13681	464	13	.	.	PUNCT
fuelectenerg-13681	465	1	120–134	120–134	NUM
fuelectenerg-13681	465	2	,	,	PUNCT
fuelectenerg-13681	465	3	2020	2020	NUM
fuelectenerg-13681	465	4	.	.	PUNCT
fuelectenerg-13681	466	1	[	[	X
fuelectenerg-13681	466	2	21	21	NUM
fuelectenerg-13681	466	3	]	]	X
fuelectenerg-13681	466	4	w.	w.	PROPN
fuelectenerg-13681	466	5	xia	xia	PROPN
fuelectenerg-13681	466	6	,	,	PUNCT
fuelectenerg-13681	466	7	g.	g.	PROPN
fuelectenerg-13681	466	8	zheng	zheng	PROPN
fuelectenerg-13681	466	9	,	,	PUNCT
fuelectenerg-13681	466	10	y.	y.	PROPN
fuelectenerg-13681	466	11	zhu	zhu	PROPN
fuelectenerg-13681	466	12	,	,	PUNCT
fuelectenerg-13681	466	13	j.	j.	PROPN
fuelectenerg-13681	466	14	zhang	zhang	PROPN
fuelectenerg-13681	466	15	,	,	PUNCT
fuelectenerg-13681	466	16	j.	j.	PROPN
fuelectenerg-13681	466	17	wang	wang	PROPN
fuelectenerg-13681	466	18	and	and	CCONJ
fuelectenerg-13681	466	19	a.	a.	PROPN
fuelectenerg-13681	466	20	p.	p.	NOUN
fuelectenerg-13681	466	21	petropulu	petropulu	VERB
fuelectenerg-13681	466	22	,	,	PUNCT
fuelectenerg-13681	466	23	"	"	PUNCT
fuelectenerg-13681	466	24	a	a	DET
fuelectenerg-13681	466	25	deep	deep	ADJ
fuelectenerg-13681	466	26	learning	learning	NOUN
fuelectenerg-13681	466	27	framework	framework	NOUN
fuelectenerg-13681	466	28	for	for	ADP
fuelectenerg-13681	466	29	optimization	optimization	NOUN
fuelectenerg-13681	466	30	of	of	ADP
fuelectenerg-13681	466	31	miso	miso	NOUN
fuelectenerg-13681	466	32	downlink	downlink	NOUN
fuelectenerg-13681	466	33	beamforming	beamforming	NOUN
fuelectenerg-13681	466	34	,	,	PUNCT
fuelectenerg-13681	466	35	"	"	PUNCT
fuelectenerg-13681	466	36	ieee	ieee	NOUN
fuelectenerg-13681	466	37	trans	tran	NOUN
fuelectenerg-13681	466	38	.	.	PUNCT
fuelectenerg-13681	467	1	commun	commun	PROPN
fuelectenerg-13681	467	2	.	.	PROPN
fuelectenerg-13681	467	3	,	,	PUNCT
fuelectenerg-13681	467	4	vol	vol	NOUN
fuelectenerg-13681	467	5	.	.	PROPN
fuelectenerg-13681	468	1	68	68	NUM
fuelectenerg-13681	468	2	,	,	PUNCT
fuelectenerg-13681	468	3	no	no	INTJ
fuelectenerg-13681	468	4	.	.	NOUN
fuelectenerg-13681	468	5	3	3	NUM
fuelectenerg-13681	468	6	,	,	PUNCT
fuelectenerg-13681	468	7	pp	pp	ADJ
fuelectenerg-13681	468	8	.	.	PUNCT
fuelectenerg-13681	469	1	1866–1880	1866–1880	NUM
fuelectenerg-13681	469	2	,	,	PUNCT
fuelectenerg-13681	469	3	2020	2020	NUM
fuelectenerg-13681	469	4	.	.	PUNCT
fuelectenerg-13681	470	1	[	[	X
fuelectenerg-13681	470	2	22	22	NUM
fuelectenerg-13681	470	3	]	]	X
fuelectenerg-13681	470	4	d.	d.	PROPN
fuelectenerg-13681	470	5	wu	wu	PROPN
fuelectenerg-13681	470	6	,	,	PUNCT
fuelectenerg-13681	470	7	y.	y.	PROPN
fuelectenerg-13681	470	8	gu	gu	PROPN
fuelectenerg-13681	470	9	,	,	PUNCT
fuelectenerg-13681	470	10	s.	s.	PROPN
fuelectenerg-13681	470	11	ma	ma	PROPN
fuelectenerg-13681	470	12	,	,	PUNCT
fuelectenerg-13681	470	13	q.	q.	PROPN
fuelectenerg-13681	470	14	li	li	PROPN
fuelectenerg-13681	470	15	,	,	PUNCT
fuelectenerg-13681	470	16	j.	j.	PROPN
fuelectenerg-13681	470	17	zhu	zhu	PROPN
fuelectenerg-13681	470	18	and	and	CCONJ
fuelectenerg-13681	470	19	w.	w.	PROPN
fuelectenerg-13681	470	20	zhang	zhang	PROPN
fuelectenerg-13681	470	21	,	,	PUNCT
fuelectenerg-13681	470	22	"	"	PUNCT
fuelectenerg-13681	470	23	distributed	distribute	VERB
fuelectenerg-13681	470	24	energy	energy	NOUN
fuelectenerg-13681	470	25	-	-	PUNCT
fuelectenerg-13681	470	26	efficient	efficient	ADJ
fuelectenerg-13681	470	27	power	power	NOUN
fuelectenerg-13681	470	28	control	control	NOUN
fuelectenerg-13681	470	29	for	for	ADP
fuelectenerg-13681	470	30	wireless	wireless	ADJ
fuelectenerg-13681	470	31	networks	network	NOUN
fuelectenerg-13681	470	32	using	use	VERB
fuelectenerg-13681	470	33	deep	deep	ADJ
fuelectenerg-13681	470	34	reinforcement	reinforcement	NOUN
fuelectenerg-13681	470	35	learning	learning	NOUN
fuelectenerg-13681	470	36	,	,	PUNCT
fuelectenerg-13681	470	37	"	"	PUNCT
fuelectenerg-13681	470	38	ieee	ieee	NOUN
fuelectenerg-13681	470	39	trans	trans	PROPN
fuelectenerg-13681	470	40	.	.	PUNCT
fuelectenerg-13681	471	1	veh	veh	PROPN
fuelectenerg-13681	471	2	.	.	PROPN
fuelectenerg-13681	471	3	technol	technol	PROPN
fuelectenerg-13681	471	4	.	.	PROPN
fuelectenerg-13681	471	5	,	,	PUNCT
fuelectenerg-13681	471	6	vol	vol	NOUN
fuelectenerg-13681	471	7	.	.	PROPN
fuelectenerg-13681	472	1	71	71	NUM
fuelectenerg-13681	472	2	,	,	PUNCT
fuelectenerg-13681	472	3	no	no	INTJ
fuelectenerg-13681	472	4	.	.	NOUN
fuelectenerg-13681	472	5	1	1	NUM
fuelectenerg-13681	472	6	,	,	PUNCT
fuelectenerg-13681	472	7	pp	pp	ADJ
fuelectenerg-13681	472	8	.	.	PUNCT
fuelectenerg-13681	473	1	1013–1017	1013–1017	NUM
fuelectenerg-13681	473	2	,	,	PUNCT
fuelectenerg-13681	473	3	2022	2022	NUM
fuelectenerg-13681	473	4	.	.	PUNCT
fuelectenerg-13681	474	1	[	[	X
fuelectenerg-13681	474	2	23	23	NUM
fuelectenerg-13681	474	3	]	]	X
fuelectenerg-13681	474	4	h.	h.	PROPN
fuelectenerg-13681	474	5	guo	guo	PROPN
fuelectenerg-13681	474	6	,	,	PUNCT
fuelectenerg-13681	474	7	y.-c	y.-c	PROPN
fuelectenerg-13681	474	8	.	.	PUNCT
fuelectenerg-13681	475	1	liang	liang	PROPN
fuelectenerg-13681	475	2	,	,	PUNCT
fuelectenerg-13681	475	3	j.	j.	PROPN
fuelectenerg-13681	475	4	chen	chen	PROPN
fuelectenerg-13681	475	5	and	and	CCONJ
fuelectenerg-13681	475	6	e.	e.	PROPN
fuelectenerg-13681	475	7	g.	g.	PROPN
fuelectenerg-13681	475	8	larsson	larsson	PROPN
fuelectenerg-13681	475	9	,	,	PUNCT
fuelectenerg-13681	475	10	"	"	PUNCT
fuelectenerg-13681	475	11	weighted	weight	VERB
fuelectenerg-13681	475	12	sum	sum	NOUN
fuelectenerg-13681	475	13	-	-	PUNCT
fuelectenerg-13681	475	14	rate	rate	NOUN
fuelectenerg-13681	475	15	maximization	maximization	NOUN
fuelectenerg-13681	475	16	for	for	ADP
fuelectenerg-13681	475	17	reconfigurable	reconfigurable	ADJ
fuelectenerg-13681	475	18	intelligent	intelligent	ADJ
fuelectenerg-13681	475	19	surface	surface	NOUN
fuelectenerg-13681	475	20	aided	aid	VERB
fuelectenerg-13681	475	21	wireless	wireless	ADJ
fuelectenerg-13681	475	22	networks	network	NOUN
fuelectenerg-13681	475	23	,	,	PUNCT
fuelectenerg-13681	475	24	"	"	PUNCT
fuelectenerg-13681	475	25	ieee	ieee	NOUN
fuelectenerg-13681	475	26	trans	tran	NOUN
fuelectenerg-13681	475	27	.	.	PUNCT
fuelectenerg-13681	476	1	wirel	wirel	PROPN
fuelectenerg-13681	476	2	.	.	PUNCT
fuelectenerg-13681	477	1	commun	commun	PROPN
fuelectenerg-13681	477	2	.	.	PROPN
fuelectenerg-13681	477	3	,	,	PUNCT
fuelectenerg-13681	477	4	vol	vol	NOUN
fuelectenerg-13681	477	5	.	.	PROPN
fuelectenerg-13681	477	6	19	19	NUM
fuelectenerg-13681	477	7	,	,	PUNCT
fuelectenerg-13681	477	8	no	no	INTJ
fuelectenerg-13681	477	9	.	.	NOUN
fuelectenerg-13681	477	10	5	5	NUM
fuelectenerg-13681	477	11	,	,	PUNCT
fuelectenerg-13681	477	12	pp	pp	ADJ
fuelectenerg-13681	477	13	.	.	PUNCT
fuelectenerg-13681	478	1	3064	3064	NUM
fuelectenerg-13681	478	2	–	–	PUNCT
fuelectenerg-13681	478	3	3076	3076	NUM
fuelectenerg-13681	478	4	,	,	PUNCT
fuelectenerg-13681	478	5	2020	2020	NUM
fuelectenerg-13681	478	6	.	.	PUNCT
fuelectenerg-13681	479	1	deep	deep	ADJ
fuelectenerg-13681	479	2	learning	learning	NOUN
fuelectenerg-13681	479	3	and	and	CCONJ
fuelectenerg-13681	479	4	traditional	traditional	ADJ
fuelectenerg-13681	479	5	algorithms	algorithm	NOUN
fuelectenerg-13681	479	6	for	for	ADP
fuelectenerg-13681	479	7	mimo	mimo	PROPN
fuelectenerg-13681	479	8	beamforming	beamforme	VERB
fuelectenerg-13681	479	9	217	217	NUM
fuelectenerg-13681	479	10	[	[	X
fuelectenerg-13681	479	11	24	24	NUM
fuelectenerg-13681	479	12	]	]	PUNCT
fuelectenerg-13681	479	13	a.	a.	NOUN
fuelectenerg-13681	479	14	m.	m.	NOUN
fuelectenerg-13681	479	15	elbir	elbir	NOUN
fuelectenerg-13681	479	16	and	and	CCONJ
fuelectenerg-13681	479	17	k.	k.	PROPN
fuelectenerg-13681	479	18	v.	v.	PROPN
fuelectenerg-13681	479	19	mishra	mishra	PROPN
fuelectenerg-13681	479	20	,	,	PUNCT
fuelectenerg-13681	479	21	"	"	PUNCT
fuelectenerg-13681	479	22	deep	deep	ADJ
fuelectenerg-13681	479	23	learning	learning	NOUN
fuelectenerg-13681	479	24	for	for	ADP
fuelectenerg-13681	479	25	mmwave	mmwave	NOUN
fuelectenerg-13681	479	26	beam	beam	NOUN
fuelectenerg-13681	479	27	and	and	CCONJ
fuelectenerg-13681	479	28	blockage	blockage	NOUN
fuelectenerg-13681	479	29	prediction	prediction	NOUN
fuelectenerg-13681	479	30	:	:	PUNCT
fuelectenerg-13681	479	31	applications	application	NOUN
fuelectenerg-13681	479	32	to	to	ADP
fuelectenerg-13681	479	33	6	6	NUM
fuelectenerg-13681	479	34	g	g	NOUN
fuelectenerg-13681	479	35	terahertz	terahertz	NOUN
fuelectenerg-13681	479	36	communications	communication	NOUN
fuelectenerg-13681	479	37	,	,	PUNCT
fuelectenerg-13681	479	38	"	"	PUNCT
fuelectenerg-13681	479	39	ieee	ieee	NOUN
fuelectenerg-13681	479	40	trans	tran	NOUN
fuelectenerg-13681	479	41	.	.	PUNCT
fuelectenerg-13681	480	1	wirel	wirel	PROPN
fuelectenerg-13681	480	2	.	.	PUNCT
fuelectenerg-13681	481	1	commun	commun	PROPN
fuelectenerg-13681	481	2	.	.	PROPN
fuelectenerg-13681	481	3	,	,	PUNCT
fuelectenerg-13681	481	4	vol	vol	NOUN
fuelectenerg-13681	481	5	.	.	PROPN
fuelectenerg-13681	481	6	21	21	NUM
fuelectenerg-13681	481	7	,	,	PUNCT
fuelectenerg-13681	481	8	no	no	INTJ
fuelectenerg-13681	481	9	.	.	NOUN
fuelectenerg-13681	481	10	11	11	NUM
fuelectenerg-13681	481	11	,	,	PUNCT
fuelectenerg-13681	481	12	pp	pp	PROPN
fuelectenerg-13681	481	13	.	.	PUNCT
fuelectenerg-13681	481	14	8691–8704	8691–8704	NUM
fuelectenerg-13681	481	15	,	,	PUNCT
fuelectenerg-13681	481	16	2022	2022	NUM
fuelectenerg-13681	481	17	.	.	PUNCT
fuelectenerg-13681	482	1	[	[	X
fuelectenerg-13681	482	2	25	25	NUM
fuelectenerg-13681	482	3	]	]	X
fuelectenerg-13681	482	4	y.	y.	PROPN
fuelectenerg-13681	482	5	jin	jin	PROPN
fuelectenerg-13681	482	6	,	,	PUNCT
fuelectenerg-13681	482	7	j.	j.	PROPN
fuelectenerg-13681	482	8	zhang	zhang	PROPN
fuelectenerg-13681	482	9	,	,	PUNCT
fuelectenerg-13681	482	10	s.	s.	PROPN
fuelectenerg-13681	482	11	jin	jin	PROPN
fuelectenerg-13681	482	12	and	and	CCONJ
fuelectenerg-13681	482	13	b.	b.	PROPN
fuelectenerg-13681	482	14	ai	ai	PROPN
fuelectenerg-13681	482	15	,	,	PUNCT
fuelectenerg-13681	482	16	"	"	PUNCT
fuelectenerg-13681	482	17	channel	channel	NOUN
fuelectenerg-13681	482	18	estimation	estimation	NOUN
fuelectenerg-13681	482	19	for	for	ADP
fuelectenerg-13681	482	20	cell	cell	NOUN
fuelectenerg-13681	482	21	-	-	PUNCT
fuelectenerg-13681	482	22	free	free	ADJ
fuelectenerg-13681	482	23	mmwave	mmwave	PROPN
fuelectenerg-13681	482	24	massive	massive	PROPN
fuelectenerg-13681	482	25	mimo	mimo	PROPN
fuelectenerg-13681	482	26	through	through	ADP
fuelectenerg-13681	482	27	deep	deep	ADJ
fuelectenerg-13681	482	28	learning	learning	NOUN
fuelectenerg-13681	482	29	,	,	PUNCT
fuelectenerg-13681	482	30	"	"	PUNCT
fuelectenerg-13681	482	31	ieee	ieee	NOUN
fuelectenerg-13681	482	32	trans	trans	PROPN
fuelectenerg-13681	482	33	.	.	PUNCT
fuelectenerg-13681	483	1	veh	veh	PROPN
fuelectenerg-13681	483	2	.	.	PROPN
fuelectenerg-13681	483	3	technol	technol	PROPN
fuelectenerg-13681	483	4	.	.	PROPN
fuelectenerg-13681	483	5	,	,	PUNCT
fuelectenerg-13681	483	6	vol	vol	NOUN
fuelectenerg-13681	483	7	.	.	PROPN
fuelectenerg-13681	484	1	68	68	NUM
fuelectenerg-13681	484	2	,	,	PUNCT
fuelectenerg-13681	484	3	no	no	INTJ
fuelectenerg-13681	484	4	.	.	NOUN
fuelectenerg-13681	484	5	10	10	NUM
fuelectenerg-13681	484	6	,	,	PUNCT
fuelectenerg-13681	484	7	pp	pp	ADJ
fuelectenerg-13681	484	8	.	.	PUNCT
fuelectenerg-13681	485	1	10325–10329	10325–10329	NUM
fuelectenerg-13681	485	2	,	,	PUNCT
fuelectenerg-13681	485	3	2019	2019	NUM
fuelectenerg-13681	485	4	.	.	PUNCT
fuelectenerg-13681	486	1	[	[	X
fuelectenerg-13681	486	2	26	26	NUM
fuelectenerg-13681	486	3	]	]	PUNCT
fuelectenerg-13681	486	4	z.	z.	PROPN
fuelectenerg-13681	486	5	liu	liu	PROPN
fuelectenerg-13681	486	6	,	,	PUNCT
fuelectenerg-13681	486	7	l.	l.	PROPN
fuelectenerg-13681	486	8	zhang	zhang	PROPN
fuelectenerg-13681	486	9	and	and	CCONJ
fuelectenerg-13681	486	10	z.	z.	PROPN
fuelectenerg-13681	486	11	ding	ding	PROPN
fuelectenerg-13681	486	12	,	,	PUNCT
fuelectenerg-13681	486	13	"	"	PUNCT
fuelectenerg-13681	486	14	exploiting	exploit	VERB
fuelectenerg-13681	486	15	bi	bi	ADJ
fuelectenerg-13681	486	16	-	-	ADJ
fuelectenerg-13681	486	17	directional	directional	ADJ
fuelectenerg-13681	486	18	channel	channel	NOUN
fuelectenerg-13681	486	19	reciprocity	reciprocity	NOUN
fuelectenerg-13681	486	20	in	in	ADP
fuelectenerg-13681	486	21	deep	deep	ADJ
fuelectenerg-13681	486	22	learning	learning	NOUN
fuelectenerg-13681	486	23	for	for	ADP
fuelectenerg-13681	486	24	low	low	ADJ
fuelectenerg-13681	486	25	rate	rate	NOUN
fuelectenerg-13681	486	26	massive	massive	ADJ
fuelectenerg-13681	486	27	mimo	mimo	PROPN
fuelectenerg-13681	486	28	csi	csi	PROPN
fuelectenerg-13681	486	29	feedback	feedback	NOUN
fuelectenerg-13681	486	30	,	,	PUNCT
fuelectenerg-13681	486	31	"	"	PUNCT
fuelectenerg-13681	486	32	ieee	ieee	NOUN
fuelectenerg-13681	486	33	wirel	wirel	NOUN
fuelectenerg-13681	486	34	.	.	PUNCT
fuelectenerg-13681	487	1	commun	commun	PROPN
fuelectenerg-13681	487	2	.	.	PUNCT
fuelectenerg-13681	488	1	lett	lett	PROPN
fuelectenerg-13681	488	2	.	.	PROPN
fuelectenerg-13681	488	3	,	,	PUNCT
fuelectenerg-13681	488	4	vol	vol	NOUN
fuelectenerg-13681	488	5	.	.	PROPN
fuelectenerg-13681	488	6	8	8	NUM
fuelectenerg-13681	488	7	,	,	PUNCT
fuelectenerg-13681	488	8	no	no	INTJ
fuelectenerg-13681	488	9	.	.	NOUN
fuelectenerg-13681	488	10	3	3	NUM
fuelectenerg-13681	488	11	,	,	PUNCT
fuelectenerg-13681	488	12	pp	pp	ADJ
fuelectenerg-13681	488	13	.	.	PUNCT
fuelectenerg-13681	489	1	889–892	889–892	NUM
fuelectenerg-13681	489	2	,	,	PUNCT
fuelectenerg-13681	489	3	2019	2019	NUM
fuelectenerg-13681	489	4	.	.	PUNCT
fuelectenerg-13681	490	1	[	[	X
fuelectenerg-13681	490	2	27	27	NUM
fuelectenerg-13681	490	3	]	]	X
fuelectenerg-13681	490	4	q.	q.	PROPN
fuelectenerg-13681	490	5	wu	wu	PROPN
fuelectenerg-13681	490	6	,	,	PUNCT
fuelectenerg-13681	490	7	k.	k.	PROPN
fuelectenerg-13681	490	8	xu	xu	PROPN
fuelectenerg-13681	490	9	,	,	PUNCT
fuelectenerg-13681	490	10	j.	j.	PROPN
fuelectenerg-13681	490	11	zhang	zhang	PROPN
fuelectenerg-13681	490	12	and	and	CCONJ
fuelectenerg-13681	490	13	p.	p.	PROPN
fuelectenerg-13681	490	14	zhang	zhang	PROPN
fuelectenerg-13681	490	15	,	,	PUNCT
fuelectenerg-13681	490	16	"	"	PUNCT
fuelectenerg-13681	490	17	a	a	DET
fuelectenerg-13681	490	18	cnn	cnn	PROPN
fuelectenerg-13681	490	19	-	-	PUNCT
fuelectenerg-13681	490	20	based	base	VERB
fuelectenerg-13681	490	21	end	end	NOUN
fuelectenerg-13681	490	22	-	-	PUNCT
fuelectenerg-13681	490	23	to	to	ADP
fuelectenerg-13681	490	24	-	-	PUNCT
fuelectenerg-13681	490	25	end	end	NOUN
fuelectenerg-13681	490	26	learning	learn	VERB
fuelectenerg-13681	490	27	framework	framework	NOUN
fuelectenerg-13681	490	28	toward	toward	ADP
fuelectenerg-13681	490	29	intelligent	intelligent	ADJ
fuelectenerg-13681	490	30	communication	communication	NOUN
fuelectenerg-13681	490	31	systems	system	NOUN
fuelectenerg-13681	490	32	,	,	PUNCT
fuelectenerg-13681	490	33	"	"	PUNCT
fuelectenerg-13681	490	34	ieee	ieee	NOUN
fuelectenerg-13681	490	35	access	access	NOUN
fuelectenerg-13681	490	36	,	,	PUNCT
fuelectenerg-13681	490	37	vol	vol	NOUN
fuelectenerg-13681	490	38	.	.	PROPN
fuelectenerg-13681	490	39	8	8	NUM
fuelectenerg-13681	490	40	,	,	PUNCT
fuelectenerg-13681	490	41	pp	pp	ADJ
fuelectenerg-13681	490	42	.	.	PUNCT
fuelectenerg-13681	491	1	77891–77904	77891–77904	NUM
fuelectenerg-13681	491	2	,	,	PUNCT
fuelectenerg-13681	491	3	2020	2020	NUM
fuelectenerg-13681	491	4	.	.	PUNCT
fuelectenerg-13681	492	1	[	[	X
fuelectenerg-13681	492	2	28	28	NUM
fuelectenerg-13681	492	3	]	]	PUNCT
fuelectenerg-13681	492	4	j.	j.	PROPN
fuelectenerg-13681	492	5	gao	gao	PROPN
fuelectenerg-13681	492	6	,	,	PUNCT
fuelectenerg-13681	492	7	c.	c.	PROPN
fuelectenerg-13681	492	8	zhong	zhong	PROPN
fuelectenerg-13681	492	9	,	,	PUNCT
fuelectenerg-13681	492	10	x.	x.	PROPN
fuelectenerg-13681	492	11	chen	chen	PROPN
fuelectenerg-13681	492	12	,	,	PUNCT
fuelectenerg-13681	492	13	h.	h.	PROPN
fuelectenerg-13681	492	14	lin	lin	PROPN
fuelectenerg-13681	492	15	and	and	CCONJ
fuelectenerg-13681	492	16	z.	z.	PROPN
fuelectenerg-13681	492	17	zhang	zhang	PROPN
fuelectenerg-13681	492	18	,	,	PUNCT
fuelectenerg-13681	492	19	"	"	PUNCT
fuelectenerg-13681	492	20	unsupervised	unsupervised	ADJ
fuelectenerg-13681	492	21	learning	learning	NOUN
fuelectenerg-13681	492	22	for	for	ADP
fuelectenerg-13681	492	23	passive	passive	ADJ
fuelectenerg-13681	492	24	beamforming	beamforming	NOUN
fuelectenerg-13681	492	25	,	,	PUNCT
fuelectenerg-13681	492	26	"	"	PUNCT
fuelectenerg-13681	492	27	ieee	ieee	X
fuelectenerg-13681	492	28	commun	commun	PROPN
fuelectenerg-13681	492	29	.	.	PUNCT
fuelectenerg-13681	493	1	lett	lett	PROPN
fuelectenerg-13681	493	2	.	.	PROPN
fuelectenerg-13681	493	3	,	,	PUNCT
fuelectenerg-13681	493	4	vol	vol	NOUN
fuelectenerg-13681	493	5	.	.	PROPN
fuelectenerg-13681	494	1	24	24	NUM
fuelectenerg-13681	494	2	,	,	PUNCT
fuelectenerg-13681	494	3	no	no	INTJ
fuelectenerg-13681	494	4	.	.	NOUN
fuelectenerg-13681	494	5	5	5	NUM
fuelectenerg-13681	494	6	,	,	PUNCT
fuelectenerg-13681	494	7	pp	pp	ADJ
fuelectenerg-13681	494	8	.	.	PUNCT
fuelectenerg-13681	495	1	1052–1056	1052–1056	NUM
fuelectenerg-13681	495	2	,	,	PUNCT
fuelectenerg-13681	495	3	2020	2020	NUM
fuelectenerg-13681	495	4	.	.	PUNCT
fuelectenerg-13681	496	1	[	[	X
fuelectenerg-13681	496	2	29	29	NUM
fuelectenerg-13681	496	3	]	]	PUNCT
fuelectenerg-13681	496	4	m.	m.	PROPN
fuelectenerg-13681	496	5	k.	k.	PROPN
fuelectenerg-13681	496	6	abdel	abdel	PROPN
fuelectenerg-13681	496	7	-	-	PUNCT
fuelectenerg-13681	496	8	aziz	aziz	PROPN
fuelectenerg-13681	496	9	,	,	PUNCT
fuelectenerg-13681	496	10	s.	s.	PROPN
fuelectenerg-13681	496	11	samarakoon	samarakoon	PROPN
fuelectenerg-13681	496	12	,	,	PUNCT
fuelectenerg-13681	496	13	m.	m.	NOUN
fuelectenerg-13681	496	14	bennis	bennis	PROPN
fuelectenerg-13681	496	15	and	and	CCONJ
fuelectenerg-13681	496	16	w.	w.	PROPN
fuelectenerg-13681	496	17	saad	saad	PROPN
fuelectenerg-13681	496	18	,	,	PUNCT
fuelectenerg-13681	496	19	"	"	PUNCT
fuelectenerg-13681	496	20	ultra	ultra	ADJ
fuelectenerg-13681	496	21	-	-	ADJ
fuelectenerg-13681	496	22	reliable	reliable	ADJ
fuelectenerg-13681	496	23	and	and	CCONJ
fuelectenerg-13681	496	24	low	low	ADJ
fuelectenerg-13681	496	25	-	-	PUNCT
fuelectenerg-13681	496	26	latency	latency	NOUN
fuelectenerg-13681	496	27	vehicular	vehicular	ADJ
fuelectenerg-13681	496	28	communication	communication	NOUN
fuelectenerg-13681	496	29	:	:	PUNCT
fuelectenerg-13681	496	30	an	an	DET
fuelectenerg-13681	496	31	active	active	ADJ
fuelectenerg-13681	496	32	learning	learning	NOUN
fuelectenerg-13681	496	33	approach	approach	NOUN
fuelectenerg-13681	496	34	,	,	PUNCT
fuelectenerg-13681	496	35	"	"	PUNCT
fuelectenerg-13681	496	36	ieee	ieee	X
fuelectenerg-13681	496	37	commun	commun	PROPN
fuelectenerg-13681	496	38	.	.	PUNCT
fuelectenerg-13681	497	1	lett	lett	PROPN
fuelectenerg-13681	497	2	.	.	PROPN
fuelectenerg-13681	497	3	,	,	PUNCT
fuelectenerg-13681	497	4	vol	vol	NOUN
fuelectenerg-13681	497	5	.	.	PROPN
fuelectenerg-13681	498	1	24	24	NUM
fuelectenerg-13681	498	2	,	,	PUNCT
fuelectenerg-13681	498	3	no	no	INTJ
fuelectenerg-13681	498	4	.	.	NOUN
fuelectenerg-13681	498	5	2	2	NUM
fuelectenerg-13681	498	6	,	,	PUNCT
fuelectenerg-13681	498	7	pp	pp	ADJ
fuelectenerg-13681	498	8	.	.	PUNCT
fuelectenerg-13681	499	1	367–370	367–370	NUM
fuelectenerg-13681	499	2	,	,	PUNCT
fuelectenerg-13681	499	3	2020	2020	NUM
fuelectenerg-13681	499	4	.	.	PUNCT
fuelectenerg-13681	500	1	[	[	X
fuelectenerg-13681	500	2	30	30	NUM
fuelectenerg-13681	500	3	]	]	PUNCT
fuelectenerg-13681	500	4	l.	l.	PROPN
fuelectenerg-13681	500	5	xu	xu	PROPN
fuelectenerg-13681	500	6	and	and	CCONJ
fuelectenerg-13681	500	7	f.	f.	PROPN
fuelectenerg-13681	500	8	gao	gao	PROPN
fuelectenerg-13681	500	9	,	,	PUNCT
fuelectenerg-13681	500	10	"	"	PUNCT
fuelectenerg-13681	500	11	model	model	NOUN
fuelectenerg-13681	500	12	aided	aid	VERB
fuelectenerg-13681	500	13	deep	deep	ADJ
fuelectenerg-13681	500	14	learning	learning	NOUN
fuelectenerg-13681	500	15	based	base	VERB
fuelectenerg-13681	500	16	mimo	mimo	PROPN
fuelectenerg-13681	500	17	ofdm	ofdm	PROPN
fuelectenerg-13681	500	18	receiver	receiver	NOUN
fuelectenerg-13681	500	19	with	with	ADP
fuelectenerg-13681	500	20	nonlinear	nonlinear	ADJ
fuelectenerg-13681	500	21	power	power	NOUN
fuelectenerg-13681	500	22	amplifiers	amplifier	NOUN
fuelectenerg-13681	500	23	,	,	PUNCT
fuelectenerg-13681	500	24	"	"	PUNCT
fuelectenerg-13681	500	25	in	in	ADP
fuelectenerg-13681	500	26	proceedings	proceeding	NOUN
fuelectenerg-13681	500	27	of	of	ADP
fuelectenerg-13681	500	28	the	the	DET
fuelectenerg-13681	500	29	ieee	ieee	NOUN
fuelectenerg-13681	500	30	wirel	wirel	NOUN
fuelectenerg-13681	500	31	.	.	PUNCT
fuelectenerg-13681	501	1	commun	commun	PROPN
fuelectenerg-13681	501	2	.	.	PUNCT
fuelectenerg-13681	502	1	netw	netw	PROPN
fuelectenerg-13681	502	2	.	.	PUNCT
fuelectenerg-13681	502	3	conf	conf	PROPN
fuelectenerg-13681	502	4	.	.	PUNCT
fuelectenerg-13681	503	1	(	(	PUNCT
fuelectenerg-13681	503	2	wcnc	wcnc	ADJ
fuelectenerg-13681	503	3	)	)	PUNCT
fuelectenerg-13681	503	4	,	,	PUNCT
fuelectenerg-13681	503	5	pp	pp	ADP
fuelectenerg-13681	503	6	.	.	PUNCT
fuelectenerg-13681	504	1	1–6	1–6	NUM
fuelectenerg-13681	504	2	,	,	PUNCT
fuelectenerg-13681	504	3	2021	2021	NUM
fuelectenerg-13681	504	4	.	.	PUNCT
fuelectenerg-13681	505	1	[	[	X
fuelectenerg-13681	505	2	31	31	NUM
fuelectenerg-13681	505	3	]	]	X
fuelectenerg-13681	505	4	w.	w.	PROPN
fuelectenerg-13681	505	5	fan	fan	PROPN
fuelectenerg-13681	505	6	,	,	PUNCT
fuelectenerg-13681	505	7	j.	j.	PROPN
fuelectenerg-13681	505	8	liang	liang	PROPN
fuelectenerg-13681	505	9	,	,	PUNCT
fuelectenerg-13681	505	10	g.	g.	PROPN
fuelectenerg-13681	505	11	yu	yu	PROPN
fuelectenerg-13681	505	12	,	,	PUNCT
fuelectenerg-13681	505	13	h.	h.	PROPN
fuelectenerg-13681	505	14	c.	c.	PROPN
fuelectenerg-13681	505	15	so	so	ADV
fuelectenerg-13681	505	16	,	,	PUNCT
fuelectenerg-13681	505	17	and	and	CCONJ
fuelectenerg-13681	505	18	j.	j.	PROPN
fuelectenerg-13681	505	19	li	li	PROPN
fuelectenerg-13681	505	20	,	,	PUNCT
fuelectenerg-13681	505	21	"	"	PUNCT
fuelectenerg-13681	505	22	robust	robust	ADJ
fuelectenerg-13681	505	23	capon	capon	NOUN
fuelectenerg-13681	505	24	beamforming	beamforme	VERB
fuelectenerg-13681	505	25	via	via	ADP
fuelectenerg-13681	505	26	admm	admm	NOUN
fuelectenerg-13681	505	27	,	,	PUNCT
fuelectenerg-13681	505	28	"	"	PUNCT
fuelectenerg-13681	505	29	in	in	ADP
fuelectenerg-13681	505	30	proceedings	proceeding	NOUN
fuelectenerg-13681	505	31	of	of	ADP
fuelectenerg-13681	505	32	the	the	DET
fuelectenerg-13681	505	33	icassp	icassp	ADJ
fuelectenerg-13681	505	34	2022	2022	NUM
fuelectenerg-13681	505	35	2022	2022	NUM
fuelectenerg-13681	505	36	ieee	ieee	PROPN
fuelectenerg-13681	505	37	international	international	ADJ
fuelectenerg-13681	505	38	conference	conference	NOUN
fuelectenerg-13681	505	39	on	on	ADP
fuelectenerg-13681	505	40	acoustics	acoustic	NOUN
fuelectenerg-13681	505	41	,	,	PUNCT
fuelectenerg-13681	505	42	speech	speech	NOUN
fuelectenerg-13681	505	43	and	and	CCONJ
fuelectenerg-13681	505	44	signal	signal	NOUN
fuelectenerg-13681	505	45	processing	processing	NOUN
fuelectenerg-13681	505	46	(	(	PUNCT
fuelectenerg-13681	505	47	icassp	icassp	PROPN
fuelectenerg-13681	505	48	)	)	PUNCT
fuelectenerg-13681	505	49	,	,	PUNCT
fuelectenerg-13681	505	50	pp	pp	PROPN
fuelectenerg-13681	505	51	.	.	PUNCT
fuelectenerg-13681	506	1	4345–4349	4345–4349	NUM
fuelectenerg-13681	506	2	,	,	PUNCT
fuelectenerg-13681	506	3	apr	apr	NOUN
fuelectenerg-13681	506	4	.	.	PROPN
fuelectenerg-13681	506	5	2019	2019	NUM
fuelectenerg-13681	506	6	.	.	PUNCT
fuelectenerg-13681	507	1	[	[	X
fuelectenerg-13681	507	2	32	32	NUM
fuelectenerg-13681	507	3	]	]	PUNCT
fuelectenerg-13681	507	4	a.	a.	NOUN
fuelectenerg-13681	507	5	zappone	zappone	NOUN
fuelectenerg-13681	507	6	,	,	PUNCT
fuelectenerg-13681	507	7	m.	m.	NOUN
fuelectenerg-13681	507	8	di	di	NOUN
fuelectenerg-13681	507	9	renzo	renzo	PROPN
fuelectenerg-13681	507	10	,	,	PUNCT
fuelectenerg-13681	507	11	and	and	CCONJ
fuelectenerg-13681	507	12	m.	m.	NOUN
fuelectenerg-13681	507	13	debbah	debbah	PROPN
fuelectenerg-13681	507	14	,	,	PUNCT
fuelectenerg-13681	507	15	"	"	PUNCT
fuelectenerg-13681	507	16	wireless	wireless	ADJ
fuelectenerg-13681	507	17	networks	network	NOUN
fuelectenerg-13681	507	18	design	design	NOUN
fuelectenerg-13681	507	19	in	in	ADP
fuelectenerg-13681	507	20	the	the	DET
fuelectenerg-13681	507	21	era	era	NOUN
fuelectenerg-13681	507	22	of	of	ADP
fuelectenerg-13681	507	23	deep	deep	ADJ
fuelectenerg-13681	507	24	learning	learning	NOUN
fuelectenerg-13681	507	25	:	:	PUNCT
fuelectenerg-13681	507	26	modelbased	modelbase	VERB
fuelectenerg-13681	507	27	,	,	PUNCT
fuelectenerg-13681	507	28	ai	ai	NOUN
fuelectenerg-13681	507	29	-	-	PUNCT
fuelectenerg-13681	507	30	based	base	VERB
fuelectenerg-13681	507	31	,	,	PUNCT
fuelectenerg-13681	507	32	or	or	CCONJ
fuelectenerg-13681	507	33	both	both	PRON
fuelectenerg-13681	507	34	?	?	PUNCT
fuelectenerg-13681	507	35	,	,	PUNCT
fuelectenerg-13681	507	36	"	"	PUNCT
fuelectenerg-13681	507	37	ieee	ieee	NOUN
fuelectenerg-13681	507	38	transactions	transaction	NOUN
fuelectenerg-13681	507	39	on	on	ADP
fuelectenerg-13681	507	40	communications	communication	NOUN
fuelectenerg-13681	507	41	,	,	PUNCT
fuelectenerg-13681	507	42	vol	vol	NOUN
fuelectenerg-13681	507	43	.	.	PROPN
fuelectenerg-13681	507	44	67	67	NUM
fuelectenerg-13681	507	45	,	,	PUNCT
fuelectenerg-13681	507	46	no	no	INTJ
fuelectenerg-13681	507	47	.	.	NOUN
fuelectenerg-13681	507	48	10	10	NUM
fuelectenerg-13681	507	49	,	,	PUNCT
fuelectenerg-13681	507	50	pp	pp	ADJ
fuelectenerg-13681	507	51	.	.	PUNCT
fuelectenerg-13681	508	1	7331–7376	7331–7376	NUM
fuelectenerg-13681	508	2	,	,	PUNCT
fuelectenerg-13681	508	3	jun	jun	PROPN
fuelectenerg-13681	508	4	.	.	PROPN
fuelectenerg-13681	508	5	2019	2019	NUM
fuelectenerg-13681	508	6	.	.	PUNCT
fuelectenerg-13681	509	1	[	[	X
fuelectenerg-13681	509	2	33	33	NUM
fuelectenerg-13681	509	3	]	]	PUNCT
fuelectenerg-13681	509	4	z.	z.	PROPN
fuelectenerg-13681	509	5	song	song	PROPN
fuelectenerg-13681	509	6	,	,	PUNCT
fuelectenerg-13681	509	7	j.	j.	PROPN
fuelectenerg-13681	509	8	yang	yang	PROPN
fuelectenerg-13681	509	9	,	,	PUNCT
fuelectenerg-13681	509	10	x.	x.	PROPN
fuelectenerg-13681	509	11	mei	mei	PROPN
fuelectenerg-13681	509	12	,	,	PUNCT
fuelectenerg-13681	509	13	t.	t.	PROPN
fuelectenerg-13681	509	14	tao	tao	PROPN
fuelectenerg-13681	509	15	,	,	PUNCT
fuelectenerg-13681	509	16	and	and	CCONJ
fuelectenerg-13681	509	17	m.	m.	PROPN
fuelectenerg-13681	509	18	xu	xu	PROPN
fuelectenerg-13681	509	19	,	,	PUNCT
fuelectenerg-13681	509	20	"	"	PUNCT
fuelectenerg-13681	509	21	harmonic	harmonic	ADJ
fuelectenerg-13681	509	22	current	current	ADJ
fuelectenerg-13681	509	23	suppression	suppression	NOUN
fuelectenerg-13681	509	24	method	method	NOUN
fuelectenerg-13681	509	25	with	with	ADP
fuelectenerg-13681	509	26	adaptive	adaptive	ADJ
fuelectenerg-13681	509	27	filter	filter	NOUN
fuelectenerg-13681	509	28	for	for	ADP
fuelectenerg-13681	509	29	permanent	permanent	ADJ
fuelectenerg-13681	509	30	magnet	magnet	NOUN
fuelectenerg-13681	509	31	synchronous	synchronous	ADJ
fuelectenerg-13681	509	32	motor	motor	NOUN
fuelectenerg-13681	509	33	,	,	PUNCT
fuelectenerg-13681	509	34	"	"	PUNCT
fuelectenerg-13681	509	35	international	international	ADJ
fuelectenerg-13681	509	36	journal	journal	NOUN
fuelectenerg-13681	509	37	of	of	ADP
fuelectenerg-13681	509	38	electronics	electronic	NOUN
fuelectenerg-13681	509	39	,	,	PUNCT
fuelectenerg-13681	509	40	vol	vol	NOUN
fuelectenerg-13681	509	41	.	.	PROPN
fuelectenerg-13681	509	42	108	108	NUM
fuelectenerg-13681	509	43	,	,	PUNCT
fuelectenerg-13681	509	44	no	no	INTJ
fuelectenerg-13681	509	45	.	.	NOUN
fuelectenerg-13681	509	46	6	6	NUM
fuelectenerg-13681	509	47	,	,	PUNCT
fuelectenerg-13681	509	48	pp	pp	ADJ
fuelectenerg-13681	509	49	.	.	PUNCT
fuelectenerg-13681	509	50	983	983	NUM
fuelectenerg-13681	509	51	–	–	PUNCT
fuelectenerg-13681	509	52	1013	1013	NUM
fuelectenerg-13681	509	53	,	,	PUNCT
fuelectenerg-13681	509	54	sep	sep	PROPN
fuelectenerg-13681	509	55	.	.	PUNCT
fuelectenerg-13681	509	56	2020	2020	NUM
fuelectenerg-13681	509	57	.	.	PUNCT
fuelectenerg-13681	510	1	[	[	X
fuelectenerg-13681	510	2	34	34	NUM
fuelectenerg-13681	510	3	]	]	X
fuelectenerg-13681	510	4	e.	e.	PROPN
fuelectenerg-13681	510	5	rodriguez	rodriguez	PROPN
fuelectenerg-13681	510	6	,	,	PUNCT
fuelectenerg-13681	510	7	b.	b.	PROPN
fuelectenerg-13681	510	8	otero	otero	PROPN
fuelectenerg-13681	510	9	,	,	PUNCT
fuelectenerg-13681	510	10	n.	n.	PROPN
fuelectenerg-13681	510	11	gutierrez	gutierrez	PROPN
fuelectenerg-13681	510	12	,	,	PUNCT
fuelectenerg-13681	510	13	and	and	CCONJ
fuelectenerg-13681	510	14	r.	r.	PROPN
fuelectenerg-13681	510	15	canal	canal	PROPN
fuelectenerg-13681	510	16	,	,	PUNCT
fuelectenerg-13681	510	17	"	"	PUNCT
fuelectenerg-13681	510	18	a	a	DET
fuelectenerg-13681	510	19	survey	survey	NOUN
fuelectenerg-13681	510	20	of	of	ADP
fuelectenerg-13681	510	21	deep	deep	ADJ
fuelectenerg-13681	510	22	learning	learning	NOUN
fuelectenerg-13681	510	23	techniques	technique	NOUN
fuelectenerg-13681	510	24	for	for	ADP
fuelectenerg-13681	510	25	cybersecurity	cybersecurity	NOUN
fuelectenerg-13681	510	26	in	in	ADP
fuelectenerg-13681	510	27	mobile	mobile	ADJ
fuelectenerg-13681	510	28	networks	network	NOUN
fuelectenerg-13681	510	29	,	,	PUNCT
fuelectenerg-13681	510	30	"	"	PUNCT
fuelectenerg-13681	510	31	ieee	ieee	NOUN
fuelectenerg-13681	510	32	communications	communication	NOUN
fuelectenerg-13681	510	33	surveys	survey	NOUN
fuelectenerg-13681	510	34	tutorials	tutorial	NOUN
fuelectenerg-13681	510	35	,	,	PUNCT
fuelectenerg-13681	510	36	vol	vol	NOUN
fuelectenerg-13681	510	37	.	.	PROPN
fuelectenerg-13681	510	38	23	23	NUM
fuelectenerg-13681	510	39	,	,	PUNCT
fuelectenerg-13681	510	40	no	no	INTJ
fuelectenerg-13681	510	41	.	.	NOUN
fuelectenerg-13681	510	42	3	3	NUM
fuelectenerg-13681	510	43	,	,	PUNCT
fuelectenerg-13681	510	44	pp	pp	ADJ
fuelectenerg-13681	510	45	.	.	PUNCT
fuelectenerg-13681	510	46	1920	1920	NUM
fuelectenerg-13681	510	47	–	–	PUNCT
fuelectenerg-13681	510	48	1955	1955	NUM
fuelectenerg-13681	510	49	,	,	PUNCT
fuelectenerg-13681	510	50	jan	jan	PROPN
fuelectenerg-13681	510	51	.	.	PROPN
fuelectenerg-13681	510	52	2021	2021	NUM
fuelectenerg-13681	510	53	.	.	PUNCT
fuelectenerg-13681	511	1	[	[	X
fuelectenerg-13681	511	2	35	35	NUM
fuelectenerg-13681	511	3	]	]	X
fuelectenerg-13681	511	4	o.	o.	PROPN
fuelectenerg-13681	511	5	s.	s.	PROPN
fuelectenerg-13681	511	6	falade	falade	PROPN
fuelectenerg-13681	511	7	,	,	PUNCT
fuelectenerg-13681	511	8	"	"	PUNCT
fuelectenerg-13681	511	9	deepmimo	deepmimo	VERB
fuelectenerg-13681	511	10	:	:	PUNCT
fuelectenerg-13681	511	11	a	a	DET
fuelectenerg-13681	511	12	generic	generic	ADJ
fuelectenerg-13681	511	13	deep	deep	ADJ
fuelectenerg-13681	511	14	learning	learning	NOUN
fuelectenerg-13681	511	15	dataset	dataset	VERB
fuelectenerg-13681	511	16	for	for	ADP
fuelectenerg-13681	511	17	millimeter	millimeter	NOUN
fuelectenerg-13681	511	18	wave	wave	NOUN
fuelectenerg-13681	511	19	and	and	CCONJ
fuelectenerg-13681	511	20	massive	massive	ADJ
fuelectenerg-13681	511	21	mimo	mimo	PROPN
fuelectenerg-13681	511	22	applications	application	NOUN
fuelectenerg-13681	511	23	to	to	ADP
fuelectenerg-13681	511	24	vehicular	vehicular	ADJ
fuelectenerg-13681	511	25	communications	communication	NOUN
fuelectenerg-13681	511	26	,	,	PUNCT
fuelectenerg-13681	511	27	"	"	PUNCT
fuelectenerg-13681	511	28	ssrn	ssrn	PROPN
fuelectenerg-13681	511	29	electronic	electronic	ADJ
fuelectenerg-13681	511	30	journal	journal	NOUN
fuelectenerg-13681	511	31	,	,	PUNCT
fuelectenerg-13681	511	32	jan	jan	PROPN
fuelectenerg-13681	511	33	.	.	PROPN
fuelectenerg-13681	511	34	2023	2023	NUM
fuelectenerg-13681	511	35	.	.	PUNCT
fuelectenerg-13681	512	1	[	[	X
fuelectenerg-13681	512	2	36	36	NUM
fuelectenerg-13681	512	3	]	]	X
fuelectenerg-13681	512	4	e.	e.	PROPN
fuelectenerg-13681	512	5	endovitskiy	endovitskiy	PROPN
fuelectenerg-13681	512	6	,	,	PUNCT
fuelectenerg-13681	512	7	a.	a.	NOUN
fuelectenerg-13681	512	8	kureev	kureev	PROPN
fuelectenerg-13681	512	9	,	,	PUNCT
fuelectenerg-13681	512	10	and	and	CCONJ
fuelectenerg-13681	512	11	e.	e.	PROPN
fuelectenerg-13681	512	12	khorov	khorov	PROPN
fuelectenerg-13681	512	13	,	,	PUNCT
fuelectenerg-13681	512	14	"	"	PUNCT
fuelectenerg-13681	512	15	reducing	reduce	VERB
fuelectenerg-13681	512	16	computational	computational	ADJ
fuelectenerg-13681	512	17	complexity	complexity	NOUN
fuelectenerg-13681	512	18	for	for	ADP
fuelectenerg-13681	512	19	the	the	DET
fuelectenerg-13681	512	20	3gpp	3gpp	PROPN
fuelectenerg-13681	512	21	tr	tr	VERB
fuelectenerg-13681	512	22	38.901	38.901	NUM
fuelectenerg-13681	512	23	mimo	mimo	PROPN
fuelectenerg-13681	512	24	channel	channel	PROPN
fuelectenerg-13681	512	25	model	model	NOUN
fuelectenerg-13681	512	26	,	,	PUNCT
fuelectenerg-13681	512	27	"	"	PUNCT
fuelectenerg-13681	512	28	ieee	ieee	NOUN
fuelectenerg-13681	512	29	wireless	wireless	ADJ
fuelectenerg-13681	512	30	communications	communication	NOUN
fuelectenerg-13681	512	31	letters	letter	NOUN
fuelectenerg-13681	512	32	,	,	PUNCT
fuelectenerg-13681	512	33	vol	vol	NOUN
fuelectenerg-13681	512	34	.	.	PROPN
fuelectenerg-13681	512	35	11	11	NUM
fuelectenerg-13681	512	36	,	,	PUNCT
fuelectenerg-13681	512	37	no	no	INTJ
fuelectenerg-13681	512	38	.	.	NOUN
fuelectenerg-13681	512	39	6	6	NUM
fuelectenerg-13681	512	40	,	,	PUNCT
fuelectenerg-13681	512	41	pp	pp	ADJ
fuelectenerg-13681	512	42	.	.	PUNCT
fuelectenerg-13681	512	43	1133–1136	1133–1136	NUM
fuelectenerg-13681	512	44	,	,	PUNCT
fuelectenerg-13681	512	45	mar	mar	PROPN
fuelectenerg-13681	512	46	.	.	PROPN
fuelectenerg-13681	512	47	2022	2022	NUM
fuelectenerg-13681	512	48	.	.	PUNCT
fuelectenerg-13681	513	1	[	[	X
fuelectenerg-13681	513	2	37	37	NUM
fuelectenerg-13681	513	3	]	]	PUNCT
fuelectenerg-13681	513	4	m.	m.	NOUN
fuelectenerg-13681	513	5	raissi	raissi	NOUN
fuelectenerg-13681	513	6	,	,	PUNCT
fuelectenerg-13681	513	7	p.	p.	NOUN
fuelectenerg-13681	513	8	perdikaris	perdikaris	NOUN
fuelectenerg-13681	513	9	,	,	PUNCT
fuelectenerg-13681	513	10	and	and	CCONJ
fuelectenerg-13681	513	11	g.	g.	PROPN
fuelectenerg-13681	513	12	e.	e.	PROPN
fuelectenerg-13681	513	13	karniadakis	karniadakis	PROPN
fuelectenerg-13681	513	14	,	,	PUNCT
fuelectenerg-13681	513	15	"	"	PUNCT
fuelectenerg-13681	513	16	physics	physics	NOUN
fuelectenerg-13681	513	17	-	-	PUNCT
fuelectenerg-13681	513	18	informed	inform	VERB
fuelectenerg-13681	513	19	neural	neural	ADJ
fuelectenerg-13681	513	20	networks	network	NOUN
fuelectenerg-13681	513	21	:	:	PUNCT
fuelectenerg-13681	513	22	a	a	DET
fuelectenerg-13681	513	23	deep	deep	ADJ
fuelectenerg-13681	513	24	learning	learning	NOUN
fuelectenerg-13681	513	25	framework	framework	NOUN
fuelectenerg-13681	513	26	for	for	ADP
fuelectenerg-13681	513	27	solving	solve	VERB
fuelectenerg-13681	513	28	forward	forward	ADV
fuelectenerg-13681	513	29	and	and	CCONJ
fuelectenerg-13681	513	30	inverse	inverse	NOUN
fuelectenerg-13681	513	31	problems	problem	NOUN
fuelectenerg-13681	513	32	involving	involve	VERB
fuelectenerg-13681	513	33	nonlinear	nonlinear	ADJ
fuelectenerg-13681	513	34	partial	partial	ADJ
fuelectenerg-13681	513	35	differential	differential	NOUN
fuelectenerg-13681	513	36	equations	equation	NOUN
fuelectenerg-13681	513	37	,	,	PUNCT
fuelectenerg-13681	513	38	"	"	PUNCT
fuelectenerg-13681	513	39	journal	journal	NOUN
fuelectenerg-13681	513	40	of	of	ADP
fuelectenerg-13681	513	41	computational	computational	ADJ
fuelectenerg-13681	513	42	physics	physics	NOUN
fuelectenerg-13681	513	43	,	,	PUNCT
fuelectenerg-13681	513	44	vol	vol	NOUN
fuelectenerg-13681	513	45	.	.	PROPN
fuelectenerg-13681	513	46	378	378	NUM
fuelectenerg-13681	513	47	,	,	PUNCT
fuelectenerg-13681	513	48	pp	pp	ADJ
fuelectenerg-13681	513	49	.	.	PUNCT
fuelectenerg-13681	514	1	686–707	686–707	NUM
fuelectenerg-13681	514	2	,	,	PUNCT
fuelectenerg-13681	514	3	nov	nov	PROPN
fuelectenerg-13681	514	4	.	.	PROPN
fuelectenerg-13681	514	5	2018	2018	NUM
fuelectenerg-13681	514	6	.	.	PUNCT
fuelectenerg-13681	515	1	[	[	X
fuelectenerg-13681	515	2	38	38	NUM
fuelectenerg-13681	515	3	]	]	PUNCT
fuelectenerg-13681	515	4	h.	h.	PROPN
fuelectenerg-13681	515	5	sarieddeen	sarieddeen	PROPN
fuelectenerg-13681	515	6	,	,	PUNCT
fuelectenerg-13681	515	7	n.	n.	PROPN
fuelectenerg-13681	515	8	saeed	saeed	PROPN
fuelectenerg-13681	515	9	,	,	PUNCT
fuelectenerg-13681	515	10	t.	t.	PROPN
fuelectenerg-13681	515	11	y.	y.	PROPN
fuelectenerg-13681	515	12	al	al	PROPN
fuelectenerg-13681	515	13	-	-	PUNCT
fuelectenerg-13681	515	14	naffouri	naffouri	NOUN
fuelectenerg-13681	515	15	,	,	PUNCT
fuelectenerg-13681	515	16	and	and	CCONJ
fuelectenerg-13681	515	17	m.-s	m.-	NOUN
fuelectenerg-13681	515	18	.	.	PUNCT
fuelectenerg-13681	516	1	alouini	alouini	PROPN
fuelectenerg-13681	516	2	,	,	PUNCT
fuelectenerg-13681	516	3	"	"	PUNCT
fuelectenerg-13681	516	4	next	next	ADJ
fuelectenerg-13681	516	5	generation	generation	NOUN
fuelectenerg-13681	516	6	terahertz	terahertz	NOUN
fuelectenerg-13681	516	7	communications	communication	NOUN
fuelectenerg-13681	516	8	:	:	PUNCT
fuelectenerg-13681	516	9	a	a	DET
fuelectenerg-13681	516	10	rendezvous	rendezvous	NOUN
fuelectenerg-13681	516	11	of	of	ADP
fuelectenerg-13681	516	12	sensing	sense	VERB
fuelectenerg-13681	516	13	,	,	PUNCT
fuelectenerg-13681	516	14	imaging	imaging	NOUN
fuelectenerg-13681	516	15	,	,	PUNCT
fuelectenerg-13681	516	16	and	and	CCONJ
fuelectenerg-13681	516	17	localization	localization	NOUN
fuelectenerg-13681	516	18	,	,	PUNCT
fuelectenerg-13681	516	19	"	"	PUNCT
fuelectenerg-13681	516	20	ieee	ieee	NOUN
fuelectenerg-13681	516	21	communications	communication	NOUN
fuelectenerg-13681	516	22	magazine	magazine	NOUN
fuelectenerg-13681	516	23	,	,	PUNCT
fuelectenerg-13681	516	24	vol	vol	NOUN
fuelectenerg-13681	516	25	.	.	PROPN
fuelectenerg-13681	517	1	58	58	NUM
fuelectenerg-13681	517	2	,	,	PUNCT
fuelectenerg-13681	517	3	no	no	INTJ
fuelectenerg-13681	517	4	.	.	NOUN
fuelectenerg-13681	517	5	5	5	NUM
fuelectenerg-13681	517	6	,	,	PUNCT
fuelectenerg-13681	517	7	pp	pp	ADJ
fuelectenerg-13681	517	8	.	.	PUNCT
fuelectenerg-13681	518	1	69–75	69–75	NUM
fuelectenerg-13681	518	2	,	,	PUNCT
fuelectenerg-13681	518	3	may	may	AUX
fuelectenerg-13681	518	4	2020	2020	NUM
fuelectenerg-13681	518	5	.	.	PUNCT