id	sid	tid	token	lemma	pos
fcis-6998	1	1	frontiers	frontier	NOUN
fcis-6998	1	2	in	in	ADP
fcis-6998	1	3	computing	computing	NOUN
fcis-6998	1	4	and	and	CCONJ
fcis-6998	1	5	intelligent	intelligent	ADJ
fcis-6998	1	6	systems	system	NOUN
fcis-6998	1	7	issn	issn	VERB
fcis-6998	1	8	:	:	PUNCT
fcis-6998	1	9	2832	2832	NUM
fcis-6998	1	10	-	-	SYM
fcis-6998	1	11	6024	6024	NUM
fcis-6998	1	12	|	|	NOUN
fcis-6998	1	13	vol	vol	NOUN
fcis-6998	1	14	.	.	PROPN
fcis-6998	2	1	3	3	NUM
fcis-6998	2	2	,	,	PUNCT
fcis-6998	2	3	no	no	INTJ
fcis-6998	2	4	.	.	NOUN
fcis-6998	2	5	2	2	NUM
fcis-6998	2	6	,	,	PUNCT
fcis-6998	2	7	2023	2023	NUM
fcis-6998	2	8	35	35	NUM
fcis-6998	2	9	hierarchical	hierarchical	ADJ
fcis-6998	2	10	federated	federated	ADJ
fcis-6998	2	11	learning	learning	NOUN
fcis-6998	2	12	with	with	ADP
fcis-6998	2	13	mobile	mobile	ADJ
fcis-6998	2	14	edge	edge	NOUN
fcis-6998	2	15	computing	computing	NOUN
fcis-6998	2	16	in	in	ADP
fcis-6998	2	17	the	the	DET
fcis-6998	2	18	internet	internet	NOUN
fcis-6998	2	19	of	of	ADP
fcis-6998	2	20	vehicles	vehicle	NOUN
fcis-6998	2	21	senwei	senwei	PROPN
fcis-6998	2	22	zhang	zhang	PROPN
fcis-6998	2	23	,	,	PUNCT
fcis-6998	2	24	fei	fei	PROPN
fcis-6998	2	25	li	li	PROPN
fcis-6998	3	1	*	*	PROPN
fcis-6998	3	2	,	,	PUNCT
fcis-6998	3	3	yi	yi	PROPN
fcis-6998	3	4	zhang	zhang	PROPN
fcis-6998	3	5	cyberspace	cyberspace	PROPN
fcis-6998	3	6	security	security	PROPN
fcis-6998	3	7	college	college	PROPN
fcis-6998	3	8	,	,	PUNCT
fcis-6998	3	9	chengdu	chengdu	PROPN
fcis-6998	3	10	university	university	PROPN
fcis-6998	3	11	of	of	ADP
fcis-6998	3	12	information	information	PROPN
fcis-6998	3	13	engineering	engineering	PROPN
fcis-6998	3	14	,	,	PUNCT
fcis-6998	3	15	chengdu	chengdu	PROPN
fcis-6998	3	16	,	,	PUNCT
fcis-6998	3	17	china	china	PROPN
fcis-6998	3	18	*	*	PUNCT
fcis-6998	3	19	corresponding	correspond	VERB
fcis-6998	3	20	author	author	NOUN
fcis-6998	3	21	:	:	PUNCT
fcis-6998	3	22	fei	fei	PROPN
fcis-6998	3	23	li	li	PROPN
fcis-6998	3	24	(	(	PUNCT
fcis-6998	3	25	email	email	NOUN
fcis-6998	3	26	:	:	PUNCT
fcis-6998	3	27	lifei@cuit.edu.cn	lifei@cuit.edu.cn	PROPN
fcis-6998	3	28	)	)	PUNCT
fcis-6998	3	29	abstract	abstract	NOUN
fcis-6998	3	30	:	:	PUNCT
fcis-6998	3	31	federated	federated	ADJ
fcis-6998	3	32	learning	learning	NOUN
fcis-6998	3	33	is	be	AUX
fcis-6998	3	34	a	a	DET
fcis-6998	3	35	distributed	distribute	VERB
fcis-6998	3	36	machine	machine	NOUN
fcis-6998	3	37	learning	learning	NOUN
fcis-6998	3	38	framework	framework	NOUN
fcis-6998	3	39	,	,	PUNCT
fcis-6998	3	40	which	which	PRON
fcis-6998	3	41	can	can	AUX
fcis-6998	3	42	be	be	AUX
fcis-6998	3	43	used	use	VERB
fcis-6998	3	44	in	in	ADP
fcis-6998	3	45	the	the	DET
fcis-6998	3	46	internet	internet	NOUN
fcis-6998	3	47	of	of	ADP
fcis-6998	3	48	vehicles	vehicle	NOUN
fcis-6998	3	49	to	to	PART
fcis-6998	3	50	train	train	VERB
fcis-6998	3	51	deep	deep	ADJ
fcis-6998	3	52	learning	learning	NOUN
fcis-6998	3	53	models	model	NOUN
fcis-6998	3	54	without	without	ADP
fcis-6998	3	55	directly	directly	ADV
fcis-6998	3	56	accessing	access	VERB
fcis-6998	3	57	the	the	DET
fcis-6998	3	58	original	original	ADJ
fcis-6998	3	59	data	datum	NOUN
fcis-6998	3	60	of	of	ADP
fcis-6998	3	61	mobile	mobile	ADJ
fcis-6998	3	62	edge	edge	NOUN
fcis-6998	3	63	vehicle	vehicle	NOUN
fcis-6998	3	64	nodes	node	NOUN
fcis-6998	3	65	.	.	PUNCT
fcis-6998	4	1	ecs	ecs	PROPN
fcis-6998	4	2	can	can	AUX
fcis-6998	4	3	access	access	VERB
fcis-6998	4	4	massive	massive	ADJ
fcis-6998	4	5	data	datum	NOUN
fcis-6998	4	6	,	,	PUNCT
fcis-6998	4	7	but	but	CCONJ
fcis-6998	4	8	it	it	PRON
fcis-6998	4	9	has	have	VERB
fcis-6998	4	10	the	the	DET
fcis-6998	4	11	characteristics	characteristic	NOUN
fcis-6998	4	12	of	of	ADP
fcis-6998	4	13	high	high	ADJ
fcis-6998	4	14	latency	latency	NOUN
fcis-6998	4	15	and	and	CCONJ
fcis-6998	4	16	high	high	ADJ
fcis-6998	4	17	communication	communication	NOUN
fcis-6998	4	18	overhead	overhead	ADV
fcis-6998	4	19	.	.	PUNCT
fcis-6998	5	1	however	however	ADV
fcis-6998	5	2	,	,	PUNCT
fcis-6998	5	3	mobile	mobile	ADJ
fcis-6998	5	4	edge	edge	NOUN
fcis-6998	5	5	computing	computing	NOUN
fcis-6998	5	6	(	(	PUNCT
fcis-6998	5	7	mec	mec	NOUN
fcis-6998	5	8	)	)	PUNCT
fcis-6998	5	9	platform	platform	NOUN
fcis-6998	5	10	can	can	AUX
fcis-6998	5	11	directly	directly	ADV
fcis-6998	5	12	and	and	CCONJ
fcis-6998	5	13	efficiently	efficiently	ADV
fcis-6998	5	14	communicate	communicate	VERB
fcis-6998	5	15	with	with	ADP
fcis-6998	5	16	mobile	mobile	ADJ
fcis-6998	5	17	edge	edge	NOUN
fcis-6998	5	18	vehicle	vehicle	NOUN
fcis-6998	5	19	nodes	node	NOUN
fcis-6998	5	20	.	.	PUNCT
fcis-6998	6	1	combining	combine	VERB
fcis-6998	6	2	the	the	DET
fcis-6998	6	3	advantages	advantage	NOUN
fcis-6998	6	4	of	of	ADP
fcis-6998	6	5	the	the	DET
fcis-6998	6	6	two	two	NUM
fcis-6998	6	7	,	,	PUNCT
fcis-6998	6	8	a	a	DET
fcis-6998	6	9	three	three	NUM
fcis-6998	6	10	-	-	PUNCT
fcis-6998	6	11	layer	layer	NOUN
fcis-6998	6	12	federated	federate	VERB
fcis-6998	6	13	learning	learn	VERB
fcis-6998	6	14	system	system	NOUN
fcis-6998	6	15	of	of	ADP
fcis-6998	6	16	edge	edge	NOUN
fcis-6998	6	17	car	car	NOUN
fcis-6998	6	18	network	network	NOUN
fcis-6998	6	19	edge	edge	NOUN
fcis-6998	6	20	server	server	PROPN
fcis-6998	6	21	cloud	cloud	NOUN
fcis-6998	6	22	server	server	NOUN
fcis-6998	6	23	is	be	AUX
fcis-6998	6	24	used	use	VERB
fcis-6998	6	25	.	.	PUNCT
fcis-6998	7	1	this	this	DET
fcis-6998	7	2	system	system	NOUN
fcis-6998	7	3	is	be	AUX
fcis-6998	7	4	supported	support	VERB
fcis-6998	7	5	by	by	ADP
fcis-6998	7	6	the	the	DET
fcis-6998	7	7	hierfedprox	hierfedprox	ADJ
fcis-6998	7	8	algorithm	algorithm	NOUN
fcis-6998	7	9	and	and	CCONJ
fcis-6998	7	10	aggregates	aggregate	VERB
fcis-6998	7	11	the	the	DET
fcis-6998	7	12	model	model	NOUN
fcis-6998	7	13	output	output	NOUN
fcis-6998	7	14	of	of	ADP
fcis-6998	7	15	the	the	DET
fcis-6998	7	16	edge	edge	NOUN
fcis-6998	7	17	car	car	NOUN
fcis-6998	7	18	to	to	ADP
fcis-6998	7	19	the	the	DET
fcis-6998	7	20	edge	edge	NOUN
fcis-6998	7	21	server	server	NOUN
fcis-6998	7	22	to	to	PART
fcis-6998	7	23	improve	improve	VERB
fcis-6998	7	24	the	the	DET
fcis-6998	7	25	model	model	NOUN
fcis-6998	7	26	learning	learn	VERB
fcis-6998	7	27	efficiency	efficiency	NOUN
fcis-6998	7	28	and	and	CCONJ
fcis-6998	7	29	reduce	reduce	VERB
fcis-6998	7	30	the	the	DET
fcis-6998	7	31	global	global	ADJ
fcis-6998	7	32	communication	communication	NOUN
fcis-6998	7	33	frequency	frequency	NOUN
fcis-6998	7	34	.	.	PUNCT
fcis-6998	8	1	the	the	DET
fcis-6998	8	2	experimental	experimental	ADJ
fcis-6998	8	3	results	result	NOUN
fcis-6998	8	4	show	show	VERB
fcis-6998	8	5	that	that	SCONJ
fcis-6998	8	6	the	the	DET
fcis-6998	8	7	system	system	NOUN
fcis-6998	8	8	can	can	AUX
fcis-6998	8	9	reduce	reduce	VERB
fcis-6998	8	10	the	the	DET
fcis-6998	8	11	training	training	NOUN
fcis-6998	8	12	time	time	NOUN
fcis-6998	8	13	and	and	CCONJ
fcis-6998	8	14	improve	improve	VERB
fcis-6998	8	15	the	the	DET
fcis-6998	8	16	accuracy	accuracy	NOUN
fcis-6998	8	17	of	of	ADP
fcis-6998	8	18	the	the	DET
fcis-6998	8	19	model	model	NOUN
fcis-6998	8	20	compared	compare	VERB
fcis-6998	8	21	with	with	ADP
fcis-6998	8	22	the	the	DET
fcis-6998	8	23	federated	federated	ADJ
fcis-6998	8	24	learning	learning	NOUN
fcis-6998	8	25	without	without	ADP
fcis-6998	8	26	introducing	introduce	VERB
fcis-6998	8	27	the	the	DET
fcis-6998	8	28	edge	edge	NOUN
fcis-6998	8	29	server	server	NOUN
fcis-6998	8	30	.	.	PUNCT
fcis-6998	9	1	keywords	keyword	NOUN
fcis-6998	9	2	:	:	PUNCT
fcis-6998	9	3	federated	federated	ADJ
fcis-6998	9	4	learning	learning	NOUN
fcis-6998	9	5	;	;	PUNCT
fcis-6998	9	6	federated	federated	ADJ
fcis-6998	9	7	proximal	proximal	NOUN
fcis-6998	9	8	;	;	PUNCT
fcis-6998	9	9	mobile	mobile	ADJ
fcis-6998	9	10	edge	edge	NOUN
fcis-6998	9	11	computing	compute	VERB
fcis-6998	9	12	platform	platform	NOUN
fcis-6998	9	13	;	;	PUNCT
fcis-6998	9	14	v2x	v2x	NOUN
fcis-6998	9	15	.	.	NOUN
fcis-6998	10	1	1	1	X
fcis-6998	10	2	.	.	X
fcis-6998	10	3	introduction	introduction	NOUN
fcis-6998	10	4	with	with	ADP
fcis-6998	10	5	the	the	DET
fcis-6998	10	6	continuous	continuous	ADJ
fcis-6998	10	7	development	development	NOUN
fcis-6998	10	8	of	of	ADP
fcis-6998	10	9	the	the	DET
fcis-6998	10	10	internet	internet	NOUN
fcis-6998	10	11	of	of	ADP
fcis-6998	10	12	vehicles	vehicle	NOUN
fcis-6998	10	13	and	and	CCONJ
fcis-6998	10	14	5	5	NUM
fcis-6998	10	15	g	g	NOUN
fcis-6998	10	16	technologies	technology	NOUN
fcis-6998	10	17	,	,	PUNCT
fcis-6998	10	18	as	as	ADV
fcis-6998	10	19	well	well	ADV
fcis-6998	10	20	as	as	ADP
fcis-6998	10	21	the	the	DET
fcis-6998	10	22	continuous	continuous	ADJ
fcis-6998	10	23	increase	increase	NOUN
fcis-6998	10	24	of	of	ADP
fcis-6998	10	25	the	the	DET
fcis-6998	10	26	number	number	NOUN
fcis-6998	10	27	of	of	ADP
fcis-6998	10	28	connected	connected	ADJ
fcis-6998	10	29	vehicles	vehicle	NOUN
fcis-6998	10	30	,	,	PUNCT
fcis-6998	10	31	a	a	DET
fcis-6998	10	32	large	large	ADJ
fcis-6998	10	33	number	number	NOUN
fcis-6998	10	34	of	of	ADP
fcis-6998	10	35	edge	edge	NOUN
fcis-6998	10	36	vehicle	vehicle	NOUN
fcis-6998	10	37	nodes	node	NOUN
fcis-6998	10	38	accessed	access	VERB
fcis-6998	10	39	in	in	ADP
fcis-6998	10	40	the	the	DET
fcis-6998	10	41	network	network	NOUN
fcis-6998	10	42	have	have	AUX
fcis-6998	10	43	generated	generate	VERB
fcis-6998	10	44	large	large	ADJ
fcis-6998	10	45	-	-	PUNCT
fcis-6998	10	46	scale	scale	NOUN
fcis-6998	10	47	operation	operation	NOUN
fcis-6998	10	48	data	datum	NOUN
fcis-6998	10	49	.	.	PUNCT
fcis-6998	11	1	the	the	DET
fcis-6998	11	2	rise	rise	NOUN
fcis-6998	11	3	of	of	ADP
fcis-6998	11	4	machine	machine	NOUN
fcis-6998	11	5	learning	learning	NOUN
fcis-6998	11	6	provides	provide	VERB
fcis-6998	11	7	a	a	DET
fcis-6998	11	8	powerful	powerful	ADJ
fcis-6998	11	9	tool	tool	NOUN
fcis-6998	11	10	for	for	ADP
fcis-6998	11	11	large	large	ADJ
fcis-6998	11	12	-	-	PUNCT
fcis-6998	11	13	scale	scale	NOUN
fcis-6998	11	14	data	datum	NOUN
fcis-6998	11	15	processing	processing	NOUN
fcis-6998	11	16	and	and	CCONJ
fcis-6998	11	17	analysis	analysis	NOUN
fcis-6998	11	18	,	,	PUNCT
fcis-6998	11	19	effectively	effectively	ADV
fcis-6998	11	20	improving	improve	VERB
fcis-6998	11	21	the	the	DET
fcis-6998	11	22	user	user	NOUN
fcis-6998	11	23	experience	experience	NOUN
fcis-6998	11	24	.	.	PUNCT
fcis-6998	12	1	although	although	SCONJ
fcis-6998	12	2	machine	machine	NOUN
fcis-6998	12	3	learning	learning	NOUN
fcis-6998	12	4	has	have	VERB
fcis-6998	12	5	a	a	DET
fcis-6998	12	6	good	good	ADJ
fcis-6998	12	7	experimental	experimental	ADJ
fcis-6998	12	8	effect	effect	NOUN
fcis-6998	12	9	in	in	ADP
fcis-6998	12	10	the	the	DET
fcis-6998	12	11	field	field	NOUN
fcis-6998	12	12	of	of	ADP
fcis-6998	12	13	the	the	DET
fcis-6998	12	14	internet	internet	NOUN
fcis-6998	12	15	of	of	ADP
fcis-6998	12	16	vehicles	vehicle	NOUN
fcis-6998	12	17	,	,	PUNCT
fcis-6998	12	18	it	it	PRON
fcis-6998	12	19	still	still	ADV
fcis-6998	12	20	faces	face	VERB
fcis-6998	12	21	the	the	DET
fcis-6998	12	22	following	follow	VERB
fcis-6998	12	23	problems	problem	NOUN
fcis-6998	12	24	:	:	PUNCT
fcis-6998	12	25	1	1	X
fcis-6998	12	26	)	)	PUNCT
fcis-6998	12	27	from	from	ADP
fcis-6998	12	28	the	the	DET
fcis-6998	12	29	perspective	perspective	NOUN
fcis-6998	12	30	of	of	ADP
fcis-6998	12	31	privacy	privacy	NOUN
fcis-6998	12	32	protection	protection	NOUN
fcis-6998	12	33	,	,	PUNCT
fcis-6998	12	34	there	there	PRON
fcis-6998	12	35	is	be	VERB
fcis-6998	12	36	a	a	DET
fcis-6998	12	37	risk	risk	NOUN
fcis-6998	12	38	of	of	ADP
fcis-6998	12	39	information	information	NOUN
fcis-6998	12	40	leakage	leakage	NOUN
fcis-6998	12	41	in	in	ADP
fcis-6998	12	42	the	the	DET
fcis-6998	12	43	transmission	transmission	NOUN
fcis-6998	12	44	and	and	CCONJ
fcis-6998	12	45	storage	storage	NOUN
fcis-6998	12	46	of	of	ADP
fcis-6998	12	47	data	datum	NOUN
fcis-6998	12	48	in	in	ADP
fcis-6998	12	49	the	the	DET
fcis-6998	12	50	internet	internet	NOUN
fcis-6998	12	51	of	of	ADP
fcis-6998	12	52	vehicles	vehicle	NOUN
fcis-6998	12	53	;	;	PUNCT
fcis-6998	12	54	2	2	X
fcis-6998	12	55	)	)	PUNCT
fcis-6998	12	56	at	at	ADP
fcis-6998	12	57	the	the	DET
fcis-6998	12	58	edge	edge	NOUN
fcis-6998	12	59	vehicle	vehicle	NOUN
fcis-6998	12	60	node	node	NOUN
fcis-6998	12	61	,	,	PUNCT
fcis-6998	12	62	due	due	ADP
fcis-6998	12	63	to	to	ADP
fcis-6998	12	64	the	the	DET
fcis-6998	12	65	limited	limited	ADJ
fcis-6998	12	66	communication	communication	NOUN
fcis-6998	12	67	and	and	CCONJ
fcis-6998	12	68	computing	computing	NOUN
fcis-6998	12	69	resources	resource	NOUN
fcis-6998	12	70	,	,	PUNCT
fcis-6998	12	71	it	it	PRON
fcis-6998	12	72	is	be	AUX
fcis-6998	12	73	difficult	difficult	ADJ
fcis-6998	12	74	to	to	PART
fcis-6998	12	75	ensure	ensure	VERB
fcis-6998	12	76	the	the	DET
fcis-6998	12	77	efficiency	efficiency	NOUN
fcis-6998	12	78	of	of	ADP
fcis-6998	12	79	model	model	NOUN
fcis-6998	12	80	training	training	NOUN
fcis-6998	12	81	while	while	SCONJ
fcis-6998	12	82	quickly	quickly	ADV
fcis-6998	12	83	and	and	CCONJ
fcis-6998	12	84	effectively	effectively	ADV
fcis-6998	12	85	calculating	calculate	VERB
fcis-6998	12	86	and	and	CCONJ
fcis-6998	12	87	processing	process	VERB
fcis-6998	12	88	the	the	DET
fcis-6998	12	89	data	datum	NOUN
fcis-6998	12	90	.	.	PUNCT
fcis-6998	13	1	in	in	ADP
fcis-6998	13	2	order	order	NOUN
fcis-6998	13	3	to	to	PART
fcis-6998	13	4	solve	solve	VERB
fcis-6998	13	5	the	the	DET
fcis-6998	13	6	problem	problem	NOUN
fcis-6998	13	7	of	of	ADP
fcis-6998	13	8	privacy	privacy	NOUN
fcis-6998	13	9	protection	protection	NOUN
fcis-6998	13	10	of	of	ADP
fcis-6998	13	11	network	network	NOUN
fcis-6998	13	12	computing	compute	VERB
fcis-6998	13	13	scenarios	scenario	NOUN
fcis-6998	13	14	for	for	ADP
fcis-6998	13	15	edge	edge	NOUN
fcis-6998	13	16	vehicles	vehicle	NOUN
fcis-6998	13	17	,	,	PUNCT
fcis-6998	13	18	federated	federated	ADJ
fcis-6998	13	19	learning	learning	NOUN
fcis-6998	13	20	came	come	VERB
fcis-6998	13	21	into	into	ADP
fcis-6998	13	22	being	being	NOUN
fcis-6998	13	23	.	.	PUNCT
fcis-6998	14	1	using	use	VERB
fcis-6998	14	2	this	this	DET
fcis-6998	14	3	algorithm	algorithm	NOUN
fcis-6998	14	4	,	,	PUNCT
fcis-6998	14	5	a	a	DET
fcis-6998	14	6	shared	share	VERB
fcis-6998	14	7	model	model	NOUN
fcis-6998	14	8	is	be	AUX
fcis-6998	14	9	trained	train	VERB
fcis-6998	14	10	cooperatively	cooperatively	ADV
fcis-6998	14	11	without	without	ADP
fcis-6998	14	12	directly	directly	ADV
fcis-6998	14	13	accessing	access	VERB
fcis-6998	14	14	the	the	DET
fcis-6998	14	15	original	original	ADJ
fcis-6998	14	16	data	datum	NOUN
fcis-6998	14	17	of	of	ADP
fcis-6998	14	18	the	the	DET
fcis-6998	14	19	edge	edge	NOUN
fcis-6998	14	20	vehicle	vehicle	NOUN
fcis-6998	14	21	nodes	nod	VERB
fcis-6998	14	22	.	.	PUNCT
fcis-6998	15	1	in	in	ADP
fcis-6998	15	2	the	the	DET
fcis-6998	15	3	fedprox	fedprox	ADJ
fcis-6998	15	4	algorithm	algorithm	NOUN
fcis-6998	15	5	,	,	PUNCT
fcis-6998	15	6	the	the	DET
fcis-6998	15	7	edge	edge	NOUN
fcis-6998	15	8	car	car	NOUN
fcis-6998	15	9	node	node	NOUN
fcis-6998	15	10	downloads	download	VERB
fcis-6998	15	11	a	a	DET
fcis-6998	15	12	global	global	ADJ
fcis-6998	15	13	machine	machine	NOUN
fcis-6998	15	14	learning	learning	NOUN
fcis-6998	15	15	model	model	NOUN
fcis-6998	15	16	from	from	ADP
fcis-6998	15	17	the	the	DET
fcis-6998	15	18	ecs	ec	NOUN
fcis-6998	15	19	,	,	PUNCT
fcis-6998	15	20	performs	perform	VERB
fcis-6998	15	21	several	several	ADJ
fcis-6998	15	22	local	local	ADJ
fcis-6998	15	23	training	training	NOUN
fcis-6998	15	24	,	,	PUNCT
fcis-6998	15	25	and	and	CCONJ
fcis-6998	15	26	then	then	ADV
fcis-6998	15	27	uploads	upload	VERB
fcis-6998	15	28	the	the	DET
fcis-6998	15	29	model	model	NOUN
fcis-6998	15	30	to	to	ADP
fcis-6998	15	31	the	the	DET
fcis-6998	15	32	ecs	ecs	PROPN
fcis-6998	15	33	for	for	ADP
fcis-6998	15	34	global	global	ADJ
fcis-6998	15	35	model	model	NOUN
fcis-6998	15	36	aggregation	aggregation	NOUN
fcis-6998	15	37	.	.	PUNCT
fcis-6998	16	1	repeat	repeat	VERB
fcis-6998	16	2	the	the	DET
fcis-6998	16	3	process	process	NOUN
fcis-6998	16	4	until	until	SCONJ
fcis-6998	16	5	the	the	DET
fcis-6998	16	6	model	model	NOUN
fcis-6998	16	7	reaches	reach	VERB
fcis-6998	16	8	the	the	DET
fcis-6998	16	9	expected	expect	VERB
fcis-6998	16	10	accuracy	accuracy	NOUN
fcis-6998	16	11	.	.	PUNCT
fcis-6998	17	1	although	although	SCONJ
fcis-6998	17	2	,	,	PUNCT
fcis-6998	17	3	federal	federal	ADJ
fcis-6998	17	4	learning	learning	NOUN
fcis-6998	17	5	will	will	AUX
fcis-6998	17	6	sink	sink	VERB
fcis-6998	17	7	the	the	DET
fcis-6998	17	8	training	training	NOUN
fcis-6998	17	9	to	to	ADP
fcis-6998	17	10	the	the	DET
fcis-6998	17	11	edge	edge	NOUN
fcis-6998	17	12	car	car	NOUN
fcis-6998	17	13	node	node	NOUN
fcis-6998	17	14	,	,	PUNCT
fcis-6998	17	15	so	so	SCONJ
fcis-6998	17	16	that	that	SCONJ
fcis-6998	17	17	the	the	DET
fcis-6998	17	18	data	datum	NOUN
fcis-6998	17	19	will	will	AUX
fcis-6998	17	20	always	always	ADV
fcis-6998	17	21	remain	remain	VERB
fcis-6998	17	22	on	on	ADP
fcis-6998	17	23	the	the	DET
fcis-6998	17	24	edge	edge	NOUN
fcis-6998	17	25	car	car	NOUN
fcis-6998	17	26	node	node	NOUN
fcis-6998	17	27	,	,	PUNCT
fcis-6998	17	28	significantly	significantly	ADV
fcis-6998	17	29	improving	improve	VERB
fcis-6998	17	30	the	the	DET
fcis-6998	17	31	privacy	privacy	NOUN
fcis-6998	17	32	of	of	ADP
fcis-6998	17	33	the	the	DET
fcis-6998	17	34	edge	edge	NOUN
fcis-6998	17	35	car	car	NOUN
fcis-6998	17	36	data	datum	NOUN
fcis-6998	17	37	.	.	PUNCT
fcis-6998	18	1	however	however	ADV
fcis-6998	18	2	,	,	PUNCT
fcis-6998	18	3	its	its	PRON
fcis-6998	18	4	iterative	iterative	NOUN
fcis-6998	18	5	training	training	NOUN
fcis-6998	18	6	process	process	NOUN
fcis-6998	18	7	requires	require	VERB
fcis-6998	18	8	hundreds	hundred	NOUN
fcis-6998	18	9	to	to	PART
fcis-6998	18	10	thousands	thousand	NOUN
fcis-6998	18	11	of	of	ADP
fcis-6998	18	12	rounds	round	NOUN
fcis-6998	18	13	of	of	ADP
fcis-6998	18	14	communication	communication	NOUN
fcis-6998	18	15	to	to	PART
fcis-6998	18	16	achieve	achieve	VERB
fcis-6998	18	17	the	the	DET
fcis-6998	18	18	ideal	ideal	ADJ
fcis-6998	18	19	model	model	NOUN
fcis-6998	18	20	accuracy	accuracy	NOUN
fcis-6998	18	21	,	,	PUNCT
fcis-6998	18	22	which	which	PRON
fcis-6998	18	23	will	will	AUX
fcis-6998	18	24	lead	lead	VERB
fcis-6998	18	25	to	to	ADP
fcis-6998	18	26	unbearable	unbearable	ADJ
fcis-6998	18	27	communication	communication	NOUN
fcis-6998	18	28	delay	delay	NOUN
fcis-6998	18	29	.	.	PUNCT
fcis-6998	19	1	there	there	PRON
fcis-6998	19	2	is	be	VERB
fcis-6998	19	3	a	a	DET
fcis-6998	19	4	problem	problem	NOUN
fcis-6998	19	5	that	that	SCONJ
fcis-6998	19	6	the	the	DET
fcis-6998	19	7	model	model	NOUN
fcis-6998	19	8	is	be	AUX
fcis-6998	19	9	difficult	difficult	ADJ
fcis-6998	19	10	to	to	PART
fcis-6998	19	11	balance	balance	VERB
fcis-6998	19	12	between	between	ADP
fcis-6998	19	13	communication	communication	NOUN
fcis-6998	19	14	efficiency	efficiency	NOUN
fcis-6998	19	15	and	and	CCONJ
fcis-6998	19	16	model	model	NOUN
fcis-6998	19	17	performance	performance	NOUN
fcis-6998	19	18	.	.	PUNCT
fcis-6998	20	1	the	the	DET
fcis-6998	20	2	mobile	mobile	ADJ
fcis-6998	20	3	edge	edge	NOUN
fcis-6998	20	4	computing	computing	NOUN
fcis-6998	20	5	platform	platform	NOUN
fcis-6998	20	6	is	be	AUX
fcis-6998	20	7	introduced	introduce	VERB
fcis-6998	20	8	as	as	ADP
fcis-6998	20	9	the	the	DET
fcis-6998	20	10	intermediate	intermediate	ADJ
fcis-6998	20	11	structure	structure	NOUN
fcis-6998	20	12	to	to	PART
fcis-6998	20	13	reduce	reduce	VERB
fcis-6998	20	14	the	the	DET
fcis-6998	20	15	communication	communication	NOUN
fcis-6998	20	16	rounds	round	NOUN
fcis-6998	20	17	between	between	ADP
fcis-6998	20	18	the	the	DET
fcis-6998	20	19	edge	edge	NOUN
fcis-6998	20	20	car	car	NOUN
fcis-6998	20	21	node	node	NOUN
fcis-6998	20	22	and	and	CCONJ
fcis-6998	20	23	the	the	DET
fcis-6998	20	24	ecs	ecs	PROPN
fcis-6998	20	25	.	.	PROPN
fcis-6998	20	26	mobile	mobile	ADJ
fcis-6998	20	27	edge	edge	NOUN
fcis-6998	20	28	computing	computing	NOUN
fcis-6998	20	29	is	be	AUX
fcis-6998	20	30	characterized	characterize	VERB
fcis-6998	20	31	by	by	ADP
fcis-6998	20	32	high	high	ADJ
fcis-6998	20	33	bandwidth	bandwidth	ADJ
fcis-6998	20	34	and	and	CCONJ
fcis-6998	20	35	low	low	ADJ
fcis-6998	20	36	latency	latency	NOUN
fcis-6998	20	37	,	,	PUNCT
fcis-6998	20	38	which	which	PRON
fcis-6998	20	39	greatly	greatly	ADV
fcis-6998	20	40	reduces	reduce	VERB
fcis-6998	20	41	the	the	DET
fcis-6998	20	42	delay	delay	NOUN
fcis-6998	20	43	of	of	ADP
fcis-6998	20	44	network	network	NOUN
fcis-6998	20	45	data	data	NOUN
fcis-6998	20	46	transmission	transmission	NOUN
fcis-6998	20	47	by	by	ADP
fcis-6998	20	48	being	be	AUX
fcis-6998	20	49	close	close	ADJ
fcis-6998	20	50	to	to	PART
fcis-6998	20	51	edge	edge	VERB
fcis-6998	20	52	car	car	NOUN
fcis-6998	20	53	nodes	node	NOUN
fcis-6998	21	1	[	[	X
fcis-6998	21	2	1,2	1,2	NUM
fcis-6998	21	3	]	]	PUNCT
fcis-6998	21	4	.	.	PUNCT
fcis-6998	22	1	the	the	DET
fcis-6998	22	2	federated	federated	ADJ
fcis-6998	22	3	learning	learning	NOUN
fcis-6998	22	4	algorithm	algorithm	NOUN
fcis-6998	22	5	based	base	VERB
fcis-6998	22	6	on	on	ADP
fcis-6998	22	7	mobile	mobile	ADJ
fcis-6998	22	8	edge	edge	NOUN
fcis-6998	22	9	computing	computing	NOUN
fcis-6998	22	10	uses	use	VERB
fcis-6998	22	11	the	the	DET
fcis-6998	22	12	nearest	near	ADJ
fcis-6998	22	13	edge	edge	NOUN
fcis-6998	22	14	server	server	NOUN
fcis-6998	22	15	as	as	ADP
fcis-6998	22	16	a	a	DET
fcis-6998	22	17	parameter	parameter	NOUN
fcis-6998	22	18	server	server	NOUN
fcis-6998	22	19	,	,	PUNCT
fcis-6998	22	20	such	such	ADJ
fcis-6998	22	21	as	as	ADP
fcis-6998	22	22	a	a	DET
fcis-6998	22	23	base	base	NOUN
fcis-6998	22	24	station	station	NOUN
fcis-6998	22	25	(	(	PUNCT
fcis-6998	22	26	macro	macro	ADJ
fcis-6998	22	27	base	base	NOUN
fcis-6998	22	28	station	station	NOUN
fcis-6998	22	29	and	and	CCONJ
fcis-6998	22	30	small	small	ADJ
fcis-6998	22	31	cell	cell	NOUN
fcis-6998	22	32	base	base	NOUN
fcis-6998	22	33	station	station	NOUN
fcis-6998	22	34	)	)	PUNCT
fcis-6998	23	1	[	[	X
fcis-6998	23	2	3	3	NUM
fcis-6998	23	3	]	]	PUNCT
fcis-6998	23	4	,	,	PUNCT
fcis-6998	23	5	while	while	SCONJ
fcis-6998	23	6	the	the	DET
fcis-6998	23	7	automobile	automobile	NOUN
fcis-6998	23	8	nodes	node	NOUN
fcis-6998	23	9	within	within	ADP
fcis-6998	23	10	the	the	DET
fcis-6998	23	11	communication	communication	NOUN
fcis-6998	23	12	range	range	NOUN
fcis-6998	23	13	of	of	ADP
fcis-6998	23	14	its	its	PRON
fcis-6998	23	15	server	server	NOUN
fcis-6998	23	16	will	will	AUX
fcis-6998	23	17	cooperate	cooperate	VERB
fcis-6998	23	18	to	to	PART
fcis-6998	23	19	train	train	VERB
fcis-6998	23	20	the	the	DET
fcis-6998	23	21	model	model	NOUN
fcis-6998	23	22	,	,	PUNCT
fcis-6998	23	23	and	and	CCONJ
fcis-6998	23	24	the	the	DET
fcis-6998	23	25	calculated	calculate	VERB
fcis-6998	23	26	delay	delay	NOUN
fcis-6998	23	27	is	be	AUX
fcis-6998	23	28	equivalent	equivalent	ADJ
fcis-6998	23	29	to	to	ADP
fcis-6998	23	30	the	the	DET
fcis-6998	23	31	communication	communication	NOUN
fcis-6998	23	32	delay	delay	NOUN
fcis-6998	23	33	of	of	ADP
fcis-6998	23	34	the	the	DET
fcis-6998	23	35	edge	edge	NOUN
fcis-6998	23	36	parameter	parameter	PROPN
fcis-6998	23	37	server	server	NOUN
fcis-6998	23	38	.	.	PUNCT
fcis-6998	24	1	therefore	therefore	ADV
fcis-6998	24	2	,	,	PUNCT
fcis-6998	24	3	it	it	PRON
fcis-6998	24	4	is	be	AUX
fcis-6998	24	5	possible	possible	ADJ
fcis-6998	24	6	to	to	PART
fcis-6998	24	7	seek	seek	VERB
fcis-6998	24	8	a	a	DET
fcis-6998	24	9	better	well	ADJ
fcis-6998	24	10	balance	balance	NOUN
fcis-6998	24	11	between	between	ADP
fcis-6998	24	12	computing	computing	NOUN
fcis-6998	24	13	and	and	CCONJ
fcis-6998	24	14	communication	communication	NOUN
fcis-6998	24	15	.	.	PUNCT
fcis-6998	25	1	in	in	ADP
fcis-6998	25	2	literature	literature	NOUN
fcis-6998	25	3	[	[	X
fcis-6998	25	4	4	4	NUM
fcis-6998	25	5	]	]	PUNCT
fcis-6998	25	6	,	,	PUNCT
fcis-6998	25	7	the	the	DET
fcis-6998	25	8	author	author	NOUN
fcis-6998	25	9	introduced	introduce	VERB
fcis-6998	25	10	mobile	mobile	ADJ
fcis-6998	25	11	edge	edge	NOUN
fcis-6998	25	12	computing	compute	VERB
fcis-6998	25	13	platform	platform	NOUN
fcis-6998	25	14	into	into	ADP
fcis-6998	25	15	federated	federated	ADJ
fcis-6998	25	16	migration	migration	NOUN
fcis-6998	25	17	learning	learning	NOUN
fcis-6998	25	18	,	,	PUNCT
fcis-6998	25	19	and	and	CCONJ
fcis-6998	25	20	obtained	obtain	VERB
fcis-6998	25	21	a	a	DET
fcis-6998	25	22	personalized	personalized	ADJ
fcis-6998	25	23	detection	detection	NOUN
fcis-6998	25	24	model	model	NOUN
fcis-6998	25	25	without	without	ADP
fcis-6998	25	26	revealing	reveal	VERB
fcis-6998	25	27	data	datum	NOUN
fcis-6998	25	28	and	and	CCONJ
fcis-6998	25	29	privacy	privacy	NOUN
fcis-6998	25	30	.	.	PUNCT
fcis-6998	26	1	in	in	ADP
fcis-6998	26	2	reference	reference	NOUN
fcis-6998	26	3	[	[	X
fcis-6998	26	4	5	5	NUM
fcis-6998	26	5	]	]	PUNCT
fcis-6998	26	6	,	,	PUNCT
fcis-6998	26	7	the	the	DET
fcis-6998	26	8	author	author	NOUN
fcis-6998	26	9	proposed	propose	VERB
fcis-6998	26	10	a	a	DET
fcis-6998	26	11	federated	federated	ADJ
fcis-6998	26	12	optimization	optimization	NOUN
fcis-6998	26	13	algorithm	algorithm	NOUN
fcis-6998	26	14	based	base	VERB
fcis-6998	26	15	on	on	ADP
fcis-6998	26	16	edge	edge	NOUN
fcis-6998	26	17	computing	compute	VERB
fcis-6998	26	18	to	to	PART
fcis-6998	26	19	solve	solve	VERB
fcis-6998	26	20	the	the	DET
fcis-6998	26	21	problem	problem	NOUN
fcis-6998	26	22	of	of	ADP
fcis-6998	26	23	large	large	ADJ
fcis-6998	26	24	amount	amount	NOUN
fcis-6998	26	25	of	of	ADP
fcis-6998	26	26	computation	computation	NOUN
fcis-6998	26	27	when	when	SCONJ
fcis-6998	26	28	mobile	mobile	ADJ
fcis-6998	26	29	devices	device	NOUN
fcis-6998	26	30	implement	implement	VERB
fcis-6998	26	31	the	the	DET
fcis-6998	26	32	federated	federated	ADJ
fcis-6998	26	33	average	average	ADJ
fcis-6998	26	34	algorithm	algorithm	NOUN
fcis-6998	26	35	.	.	PUNCT
fcis-6998	27	1	the	the	DET
fcis-6998	27	2	local	local	ADJ
fcis-6998	27	3	update	update	NOUN
fcis-6998	27	4	process	process	NOUN
fcis-6998	27	5	is	be	AUX
fcis-6998	27	6	divided	divide	VERB
fcis-6998	27	7	into	into	ADP
fcis-6998	27	8	mobile	mobile	ADJ
fcis-6998	27	9	edge	edge	NOUN
fcis-6998	27	10	car	car	NOUN
fcis-6998	27	11	node	node	NOUN
fcis-6998	27	12	and	and	CCONJ
fcis-6998	27	13	edge	edge	NOUN
fcis-6998	27	14	server	server	NOUN
fcis-6998	27	15	to	to	PART
fcis-6998	27	16	complete	complete	VERB
fcis-6998	27	17	,	,	PUNCT
fcis-6998	27	18	and	and	CCONJ
fcis-6998	27	19	the	the	DET
fcis-6998	27	20	global	global	ADJ
fcis-6998	27	21	aggregation	aggregation	NOUN
fcis-6998	27	22	process	process	NOUN
fcis-6998	27	23	is	be	AUX
fcis-6998	27	24	carried	carry	VERB
fcis-6998	27	25	out	out	ADP
fcis-6998	27	26	between	between	ADP
fcis-6998	27	27	the	the	DET
fcis-6998	27	28	edge	edge	NOUN
fcis-6998	27	29	server	server	NOUN
fcis-6998	27	30	and	and	CCONJ
fcis-6998	27	31	ecs	ec	NOUN
fcis-6998	27	32	to	to	PART
fcis-6998	27	33	reduce	reduce	VERB
fcis-6998	27	34	the	the	DET
fcis-6998	27	35	total	total	ADJ
fcis-6998	27	36	communication	communication	NOUN
fcis-6998	27	37	and	and	CCONJ
fcis-6998	27	38	computing	computing	NOUN
fcis-6998	27	39	costs	cost	NOUN
fcis-6998	27	40	of	of	ADP
fcis-6998	27	41	the	the	DET
fcis-6998	27	42	edge	edge	NOUN
fcis-6998	27	43	car	car	NOUN
fcis-6998	27	44	node	node	NOUN
fcis-6998	27	45	.	.	PUNCT
fcis-6998	28	1	combining	combine	VERB
fcis-6998	28	2	the	the	DET
fcis-6998	28	3	advantages	advantage	NOUN
fcis-6998	28	4	of	of	ADP
fcis-6998	28	5	federated	federated	ADJ
fcis-6998	28	6	learning	learning	NOUN
fcis-6998	28	7	and	and	CCONJ
fcis-6998	28	8	mobile	mobile	ADJ
fcis-6998	28	9	edge	edge	NOUN
fcis-6998	28	10	computing	computing	NOUN
fcis-6998	28	11	,	,	PUNCT
fcis-6998	28	12	we	we	PRON
fcis-6998	28	13	use	use	VERB
fcis-6998	28	14	a	a	DET
fcis-6998	28	15	three	three	NUM
fcis-6998	28	16	-	-	PUNCT
fcis-6998	28	17	tier	tier	NOUN
fcis-6998	28	18	architecture	architecture	NOUN
fcis-6998	28	19	of	of	ADP
fcis-6998	28	20	edge	edge	NOUN
fcis-6998	28	21	car	car	NOUN
fcis-6998	28	22	network	network	NOUN
fcis-6998	28	23	edge	edge	NOUN
fcis-6998	28	24	server	server	PROPN
fcis-6998	28	25	cloud	cloud	PROPN
fcis-6998	28	26	server	server	NOUN
fcis-6998	28	27	federated	federated	ADJ
fcis-6998	28	28	learning	learning	NOUN
fcis-6998	28	29	,	,	PUNCT
fcis-6998	28	30	whose	whose	DET
fcis-6998	28	31	algorithm	algorithm	NOUN
fcis-6998	28	32	is	be	AUX
fcis-6998	28	33	hierfedprox	hierfedprox	ADJ
fcis-6998	28	34	.	.	PUNCT
fcis-6998	29	1	compared	compare	VERB
fcis-6998	29	2	with	with	ADP
fcis-6998	29	3	hierfavg	hierfavg	PROPN
fcis-6998	29	4	algorithm	algorithm	PROPN
fcis-6998	29	5	,	,	PUNCT
fcis-6998	29	6	hierfedprox	hierfedprox	ADJ
fcis-6998	29	7	algorithm	algorithm	NOUN
fcis-6998	29	8	has	have	VERB
fcis-6998	29	9	inconsistent	inconsistent	ADJ
fcis-6998	29	10	distribution	distribution	NOUN
fcis-6998	29	11	among	among	ADP
fcis-6998	29	12	edge	edge	NOUN
fcis-6998	29	13	data	datum	NOUN
fcis-6998	29	14	,	,	PUNCT
fcis-6998	29	15	which	which	PRON
fcis-6998	29	16	can	can	AUX
fcis-6998	29	17	improve	improve	VERB
fcis-6998	29	18	the	the	DET
fcis-6998	29	19	robustness	robustness	NOUN
fcis-6998	29	20	and	and	CCONJ
fcis-6998	29	21	stability	stability	NOUN
fcis-6998	29	22	of	of	ADP
fcis-6998	29	23	model	model	NOUN
fcis-6998	29	24	convergence	convergence	NOUN
fcis-6998	29	25	.	.	PUNCT
fcis-6998	30	1	the	the	DET
fcis-6998	30	2	three	three	NUM
fcis-6998	30	3	-	-	PUNCT
fcis-6998	30	4	tier	tier	NOUN
fcis-6998	30	5	architecture	architecture	NOUN
fcis-6998	30	6	can	can	AUX
fcis-6998	30	7	use	use	VERB
fcis-6998	30	8	the	the	DET
fcis-6998	30	9	edge	edge	NOUN
fcis-6998	30	10	server	server	NOUN
fcis-6998	30	11	as	as	ADP
fcis-6998	30	12	the	the	DET
fcis-6998	30	13	intermediary	intermediary	NOUN
fcis-6998	30	14	between	between	ADP
fcis-6998	30	15	the	the	DET
fcis-6998	30	16	edge	edge	NOUN
fcis-6998	30	17	car	car	NOUN
fcis-6998	30	18	network	network	NOUN
fcis-6998	30	19	and	and	CCONJ
fcis-6998	30	20	the	the	DET
fcis-6998	30	21	cloud	cloud	NOUN
fcis-6998	30	22	server	server	NOUN
fcis-6998	30	23	,	,	PUNCT
fcis-6998	30	24	which	which	PRON
fcis-6998	30	25	can	can	AUX
fcis-6998	30	26	significantly	significantly	ADV
fcis-6998	30	27	reduce	reduce	VERB
fcis-6998	30	28	the	the	DET
fcis-6998	30	29	communication	communication	NOUN
fcis-6998	30	30	cost	cost	NOUN
fcis-6998	30	31	,	,	PUNCT
fcis-6998	30	32	while	while	SCONJ
fcis-6998	30	33	still	still	ADV
fcis-6998	30	34	using	use	VERB
fcis-6998	30	35	the	the	DET
fcis-6998	30	36	rich	rich	ADJ
fcis-6998	30	37	data	datum	NOUN
fcis-6998	30	38	of	of	ADP
fcis-6998	30	39	the	the	DET
fcis-6998	30	40	edge	edge	NOUN
fcis-6998	30	41	car	car	NOUN
fcis-6998	30	42	network	network	NOUN
fcis-6998	30	43	to	to	PART
fcis-6998	30	44	improve	improve	VERB
fcis-6998	30	45	the	the	DET
fcis-6998	30	46	machine	machine	NOUN
fcis-6998	30	47	learning	learn	VERB
fcis-6998	30	48	efficiency	efficiency	NOUN
fcis-6998	30	49	.	.	PUNCT
fcis-6998	31	1	2	2	X
fcis-6998	31	2	.	.	X
fcis-6998	31	3	research	research	NOUN
fcis-6998	31	4	status	status	NOUN
fcis-6998	31	5	2.1	2.1	NUM
fcis-6998	31	6	.	.	PUNCT
fcis-6998	32	1	mobile	mobile	ADJ
fcis-6998	32	2	edge	edge	NOUN
fcis-6998	32	3	computing	compute	VERB
fcis-6998	32	4	mobile	mobile	ADJ
fcis-6998	32	5	edge	edge	NOUN
fcis-6998	32	6	computing	computing	NOUN
fcis-6998	32	7	provides	provide	VERB
fcis-6998	32	8	internet	internet	NOUN
fcis-6998	32	9	service	service	NOUN
fcis-6998	32	10	environment	environment	NOUN
fcis-6998	32	11	and	and	CCONJ
fcis-6998	32	12	cloud	cloud	NOUN
fcis-6998	32	13	computing	computing	NOUN
fcis-6998	32	14	capability	capability	NOUN
fcis-6998	32	15	for	for	ADP
fcis-6998	32	16	5	5	NUM
fcis-6998	32	17	g	g	NOUN
fcis-6998	32	18	,	,	PUNCT
fcis-6998	32	19	wifi	wifi	NOUN
fcis-6998	32	20	and	and	CCONJ
fcis-6998	32	21	other	other	ADJ
fcis-6998	32	22	wireless	wireless	ADJ
fcis-6998	32	23	networks	network	NOUN
fcis-6998	32	24	,	,	PUNCT
fcis-6998	32	25	and	and	CCONJ
fcis-6998	32	26	can	can	AUX
fcis-6998	32	27	meet	meet	VERB
fcis-6998	32	28	the	the	DET
fcis-6998	32	29	communication	communication	NOUN
fcis-6998	32	30	36	36	NUM
fcis-6998	32	31	requirements	requirement	NOUN
fcis-6998	32	32	of	of	ADP
fcis-6998	32	33	ultra	ultra	ADJ
fcis-6998	32	34	-	-	ADJ
fcis-6998	32	35	high	high	ADJ
fcis-6998	32	36	reliability	reliability	NOUN
fcis-6998	32	37	and	and	CCONJ
fcis-6998	32	38	ultra	ultra	ADJ
fcis-6998	32	39	-	-	ADJ
fcis-6998	32	40	low	low	ADJ
fcis-6998	32	41	delay	delay	NOUN
fcis-6998	32	42	of	of	ADP
fcis-6998	32	43	5	5	NUM
fcis-6998	32	44	g	g	NOUN
fcis-6998	32	45	function	function	NOUN
fcis-6998	32	46	edge	edge	NOUN
fcis-6998	32	47	car	car	NOUN
fcis-6998	32	48	nodes	node	NOUN
fcis-6998	32	49	.	.	PUNCT
fcis-6998	33	1	in	in	ADP
fcis-6998	33	2	document	document	NOUN
fcis-6998	33	3	[	[	X
fcis-6998	33	4	6	6	NUM
fcis-6998	33	5	]	]	PUNCT
fcis-6998	33	6	,	,	PUNCT
fcis-6998	33	7	a	a	DET
fcis-6998	33	8	reliable	reliable	ADJ
fcis-6998	33	9	low	low	ADJ
fcis-6998	33	10	delay	delay	NOUN
fcis-6998	33	11	vehicle	vehicle	NOUN
fcis-6998	33	12	communication	communication	NOUN
fcis-6998	33	13	architecture	architecture	NOUN
fcis-6998	33	14	based	base	VERB
fcis-6998	33	15	on	on	ADP
fcis-6998	33	16	mec	mec	PROPN
fcis-6998	33	17	(	(	PUNCT
fcis-6998	33	18	mobile	mobile	ADJ
fcis-6998	33	19	edge	edge	NOUN
fcis-6998	33	20	computing	compute	VERB
fcis-6998	33	21	platform	platform	NOUN
fcis-6998	33	22	)	)	PUNCT
fcis-6998	33	23	is	be	AUX
fcis-6998	33	24	proposed	propose	VERB
fcis-6998	33	25	,	,	PUNCT
fcis-6998	33	26	which	which	PRON
fcis-6998	33	27	supports	support	VERB
fcis-6998	33	28	seamless	seamless	ADJ
fcis-6998	33	29	handover	handover	NOUN
fcis-6998	33	30	of	of	ADP
fcis-6998	33	31	high	high	ADJ
fcis-6998	33	32	mobility	mobility	NOUN
fcis-6998	33	33	v2v	v2v	NOUN
fcis-6998	33	34	communication	communication	NOUN
fcis-6998	33	35	,	,	PUNCT
fcis-6998	33	36	and	and	CCONJ
fcis-6998	33	37	proves	prove	VERB
fcis-6998	33	38	that	that	SCONJ
fcis-6998	33	39	a	a	DET
fcis-6998	33	40	flexible	flexible	ADJ
fcis-6998	33	41	and	and	CCONJ
fcis-6998	33	42	active	active	ADJ
fcis-6998	33	43	service	service	NOUN
fcis-6998	33	44	mec	mec	PROPN
fcis-6998	33	45	selection	selection	NOUN
fcis-6998	33	46	has	have	VERB
fcis-6998	33	47	better	well	ADJ
fcis-6998	33	48	performance	performance	NOUN
fcis-6998	33	49	.	.	PUNCT
fcis-6998	34	1	for	for	ADP
fcis-6998	34	2	the	the	DET
fcis-6998	34	3	research	research	NOUN
fcis-6998	34	4	of	of	ADP
fcis-6998	34	5	mec	mec	PROPN
fcis-6998	34	6	auxiliary	auxiliary	ADJ
fcis-6998	34	7	communication	communication	NOUN
fcis-6998	34	8	,	,	PUNCT
fcis-6998	34	9	in	in	ADP
fcis-6998	34	10	literature	literature	NOUN
fcis-6998	34	11	[	[	X
fcis-6998	34	12	7	7	NUM
fcis-6998	34	13	]	]	PUNCT
fcis-6998	34	14	,	,	PUNCT
fcis-6998	34	15	considering	consider	VERB
fcis-6998	34	16	traffic	traffic	NOUN
fcis-6998	34	17	load	load	NOUN
fcis-6998	34	18	,	,	PUNCT
fcis-6998	34	19	computing	computing	NOUN
fcis-6998	34	20	and	and	CCONJ
fcis-6998	34	21	storage	storage	NOUN
fcis-6998	34	22	capacity	capacity	NOUN
fcis-6998	34	23	,	,	PUNCT
fcis-6998	34	24	energy	energy	NOUN
fcis-6998	34	25	consumption	consumption	NOUN
fcis-6998	34	26	and	and	CCONJ
fcis-6998	34	27	delay	delay	NOUN
fcis-6998	34	28	,	,	PUNCT
fcis-6998	34	29	it	it	PRON
fcis-6998	34	30	is	be	AUX
fcis-6998	34	31	proved	prove	VERB
fcis-6998	34	32	that	that	SCONJ
fcis-6998	34	33	mec	mec	PROPN
fcis-6998	34	34	can	can	AUX
fcis-6998	34	35	reduce	reduce	VERB
fcis-6998	34	36	resource	resource	NOUN
fcis-6998	34	37	leasing	leasing	NOUN
fcis-6998	34	38	.	.	PUNCT
fcis-6998	35	1	in	in	ADP
fcis-6998	35	2	the	the	DET
fcis-6998	35	3	paper	paper	NOUN
fcis-6998	35	4	[	[	X
fcis-6998	35	5	8	8	NUM
fcis-6998	35	6	]	]	PUNCT
fcis-6998	35	7	,	,	PUNCT
fcis-6998	35	8	a	a	DET
fcis-6998	35	9	new	new	ADJ
fcis-6998	35	10	mec	mec	X
fcis-6998	35	11	based	base	VERB
fcis-6998	35	12	vehicle	vehicle	NOUN
fcis-6998	35	13	network	network	NOUN
fcis-6998	35	14	framework	framework	NOUN
fcis-6998	35	15	is	be	AUX
fcis-6998	35	16	proposed	propose	VERB
fcis-6998	35	17	,	,	PUNCT
fcis-6998	35	18	a	a	DET
fcis-6998	35	19	task	task	NOUN
fcis-6998	35	20	file	file	NOUN
fcis-6998	35	21	transmission	transmission	NOUN
fcis-6998	35	22	strategy	strategy	NOUN
fcis-6998	35	23	with	with	ADP
fcis-6998	35	24	predictive	predictive	ADJ
fcis-6998	35	25	v2v	v2v	NOUN
fcis-6998	35	26	relay	relay	NOUN
fcis-6998	35	27	is	be	AUX
fcis-6998	35	28	designed	design	VERB
fcis-6998	35	29	,	,	PUNCT
fcis-6998	35	30	and	and	CCONJ
fcis-6998	35	31	an	an	DET
fcis-6998	35	32	optimal	optimal	ADJ
fcis-6998	35	33	predictive	predictive	ADJ
fcis-6998	35	34	combined	combine	VERB
fcis-6998	35	35	mode	mode	NOUN
fcis-6998	35	36	unloading	unloading	NOUN
fcis-6998	35	37	scheme	scheme	NOUN
fcis-6998	35	38	is	be	AUX
fcis-6998	35	39	proposed	propose	VERB
fcis-6998	35	40	,	,	PUNCT
fcis-6998	35	41	which	which	PRON
fcis-6998	35	42	greatly	greatly	ADV
fcis-6998	35	43	reduces	reduce	VERB
fcis-6998	35	44	the	the	DET
fcis-6998	35	45	unloading	unloading	NOUN
fcis-6998	35	46	cost	cost	NOUN
fcis-6998	35	47	.	.	PUNCT
fcis-6998	36	1	in	in	ADP
fcis-6998	36	2	order	order	NOUN
fcis-6998	36	3	to	to	PART
fcis-6998	36	4	avoid	avoid	VERB
fcis-6998	36	5	interruptions	interruption	NOUN
fcis-6998	36	6	in	in	ADP
fcis-6998	36	7	the	the	DET
fcis-6998	36	8	unloading	unloading	NOUN
fcis-6998	36	9	process	process	NOUN
fcis-6998	36	10	,	,	PUNCT
fcis-6998	36	11	literature	literature	NOUN
fcis-6998	36	12	[	[	X
fcis-6998	36	13	9	9	NUM
fcis-6998	36	14	]	]	PUNCT
fcis-6998	36	15	proposed	propose	VERB
fcis-6998	36	16	a	a	DET
fcis-6998	36	17	dynamic	dynamic	ADJ
fcis-6998	36	18	unloading	unloading	NOUN
fcis-6998	36	19	scheduling	scheduling	NOUN
fcis-6998	36	20	scheme	scheme	NOUN
fcis-6998	36	21	based	base	VERB
fcis-6998	36	22	on	on	ADP
fcis-6998	36	23	mec	mec	PROPN
fcis-6998	36	24	for	for	ADP
fcis-6998	36	25	the	the	DET
fcis-6998	36	26	vehicle	vehicle	NOUN
fcis-6998	36	27	network	network	NOUN
fcis-6998	36	28	,	,	PUNCT
fcis-6998	36	29	which	which	PRON
fcis-6998	36	30	divided	divide	VERB
fcis-6998	36	31	the	the	DET
fcis-6998	36	32	entire	entire	ADJ
fcis-6998	36	33	unloading	unloading	NOUN
fcis-6998	36	34	task	task	NOUN
fcis-6998	36	35	into	into	ADP
fcis-6998	36	36	small	small	ADJ
fcis-6998	36	37	task	task	NOUN
fcis-6998	36	38	units	unit	NOUN
fcis-6998	36	39	,	,	PUNCT
fcis-6998	36	40	and	and	CCONJ
fcis-6998	36	41	then	then	ADV
fcis-6998	36	42	calculated	calculate	VERB
fcis-6998	36	43	the	the	DET
fcis-6998	36	44	number	number	NOUN
fcis-6998	36	45	of	of	ADP
fcis-6998	36	46	units	unit	NOUN
fcis-6998	36	47	allocated	allocate	VERB
fcis-6998	36	48	and	and	CCONJ
fcis-6998	36	49	the	the	DET
fcis-6998	36	50	optimal	optimal	ADJ
fcis-6998	36	51	unloading	unloading	NOUN
fcis-6998	36	52	ratio	ratio	NOUN
fcis-6998	36	53	according	accord	VERB
fcis-6998	36	54	to	to	ADP
fcis-6998	36	55	different	different	ADJ
fcis-6998	36	56	constraints	constraint	NOUN
fcis-6998	36	57	.	.	PUNCT
fcis-6998	37	1	with	with	ADP
fcis-6998	37	2	this	this	DET
fcis-6998	37	3	scheduling	scheduling	NOUN
fcis-6998	37	4	scheme	scheme	NOUN
fcis-6998	37	5	,	,	PUNCT
fcis-6998	37	6	the	the	DET
fcis-6998	37	7	delay	delay	NOUN
fcis-6998	37	8	in	in	ADP
fcis-6998	37	9	each	each	DET
fcis-6998	37	10	cell	cell	NOUN
fcis-6998	37	11	reaches	reach	VERB
fcis-6998	37	12	a	a	DET
fcis-6998	37	13	minimum	minimum	ADJ
fcis-6998	37	14	value	value	NOUN
fcis-6998	37	15	.	.	PUNCT
fcis-6998	38	1	the	the	DET
fcis-6998	38	2	dynamic	dynamic	ADJ
fcis-6998	38	3	unloading	unloading	NOUN
fcis-6998	38	4	scheduling	scheduling	NOUN
fcis-6998	38	5	scheme	scheme	NOUN
fcis-6998	38	6	does	do	AUX
fcis-6998	38	7	not	not	PART
fcis-6998	38	8	interrupt	interrupt	VERB
fcis-6998	38	9	the	the	DET
fcis-6998	38	10	unloading	unloading	NOUN
fcis-6998	38	11	process	process	NOUN
fcis-6998	38	12	.	.	PUNCT
fcis-6998	39	1	mec	mec	PROPN
fcis-6998	39	2	has	have	AUX
fcis-6998	39	3	been	be	AUX
fcis-6998	39	4	widely	widely	ADV
fcis-6998	39	5	used	use	VERB
fcis-6998	39	6	in	in	ADP
fcis-6998	39	7	the	the	DET
fcis-6998	39	8	internet	internet	NOUN
fcis-6998	39	9	of	of	ADP
fcis-6998	39	10	vehicles	vehicle	NOUN
fcis-6998	39	11	to	to	PART
fcis-6998	39	12	complete	complete	VERB
fcis-6998	39	13	task	task	PROPN
fcis-6998	39	14	unloading	unloading	NOUN
fcis-6998	39	15	,	,	PUNCT
fcis-6998	39	16	data	datum	NOUN
fcis-6998	39	17	sharing	sharing	NOUN
fcis-6998	39	18	and	and	CCONJ
fcis-6998	39	19	content	content	NOUN
fcis-6998	39	20	caching	cache	VERB
fcis-6998	39	21	of	of	ADP
fcis-6998	39	22	the	the	DET
fcis-6998	39	23	internet	internet	NOUN
fcis-6998	39	24	of	of	ADP
fcis-6998	39	25	vehicles	vehicle	NOUN
fcis-6998	39	26	to	to	PART
fcis-6998	39	27	meet	meet	VERB
fcis-6998	39	28	the	the	DET
fcis-6998	39	29	growing	grow	VERB
fcis-6998	39	30	demand	demand	NOUN
fcis-6998	39	31	for	for	ADP
fcis-6998	39	32	computing	compute	VERB
fcis-6998	39	33	intensive	intensive	ADJ
fcis-6998	39	34	and	and	CCONJ
fcis-6998	39	35	delay	delay	VERB
fcis-6998	39	36	sensitive	sensitive	ADJ
fcis-6998	39	37	automotive	automotive	ADJ
fcis-6998	39	38	applications	application	NOUN
fcis-6998	39	39	.	.	PUNCT
fcis-6998	40	1	2.2	2.2	NUM
fcis-6998	40	2	.	.	PUNCT
fcis-6998	41	1	federal	federal	ADJ
fcis-6998	41	2	learning	learning	PROPN
fcis-6998	41	3	mcmahan	mcmahan	PROPN
fcis-6998	42	1	[	[	X
fcis-6998	42	2	10	10	NUM
fcis-6998	42	3	]	]	PUNCT
fcis-6998	42	4	first	first	ADV
fcis-6998	42	5	proposed	propose	VERB
fcis-6998	42	6	federated	federated	ADJ
fcis-6998	42	7	learning	learning	NOUN
fcis-6998	42	8	technology	technology	NOUN
fcis-6998	42	9	in	in	ADP
fcis-6998	42	10	2016	2016	NUM
fcis-6998	42	11	.	.	PUNCT
fcis-6998	43	1	it	it	PRON
fcis-6998	43	2	is	be	AUX
fcis-6998	43	3	essentially	essentially	ADV
fcis-6998	43	4	a	a	DET
fcis-6998	43	5	distributed	distribute	VERB
fcis-6998	43	6	machine	machine	NOUN
fcis-6998	43	7	learning	learning	NOUN
fcis-6998	43	8	framework	framework	NOUN
fcis-6998	43	9	,	,	PUNCT
fcis-6998	43	10	which	which	PRON
fcis-6998	43	11	can	can	AUX
fcis-6998	43	12	not	not	PART
fcis-6998	43	13	only	only	ADV
fcis-6998	43	14	protect	protect	VERB
fcis-6998	43	15	user	user	NOUN
fcis-6998	43	16	privacy	privacy	NOUN
fcis-6998	43	17	in	in	ADP
fcis-6998	43	18	the	the	DET
fcis-6998	43	19	process	process	NOUN
fcis-6998	43	20	of	of	ADP
fcis-6998	43	21	machine	machine	NOUN
fcis-6998	43	22	learning	learning	NOUN
fcis-6998	43	23	,	,	PUNCT
fcis-6998	43	24	but	but	CCONJ
fcis-6998	43	25	also	also	ADV
fcis-6998	43	26	realize	realize	VERB
fcis-6998	43	27	data	datum	NOUN
fcis-6998	43	28	sharing	sharing	NOUN
fcis-6998	43	29	in	in	ADP
fcis-6998	43	30	the	the	DET
fcis-6998	43	31	case	case	NOUN
fcis-6998	43	32	of	of	ADP
fcis-6998	43	33	raw	raw	ADJ
fcis-6998	43	34	data	data	NOUN
fcis-6998	43	35	aggregation	aggregation	NOUN
fcis-6998	43	36	.	.	PUNCT
fcis-6998	44	1	according	accord	VERB
fcis-6998	44	2	to	to	ADP
fcis-6998	44	3	the	the	DET
fcis-6998	44	4	characteristics	characteristic	NOUN
fcis-6998	44	5	of	of	ADP
fcis-6998	44	6	the	the	DET
fcis-6998	44	7	customer	customer	NOUN
fcis-6998	44	8	data	datum	NOUN
fcis-6998	44	9	set	set	VERB
fcis-6998	44	10	participating	participate	VERB
fcis-6998	44	11	in	in	ADP
fcis-6998	44	12	the	the	DET
fcis-6998	44	13	training	training	NOUN
fcis-6998	44	14	,	,	PUNCT
fcis-6998	44	15	federal	federal	ADJ
fcis-6998	44	16	learning	learning	NOUN
fcis-6998	44	17	can	can	AUX
fcis-6998	44	18	be	be	AUX
fcis-6998	44	19	divided	divide	VERB
fcis-6998	44	20	into	into	ADP
fcis-6998	44	21	horizontal	horizontal	ADJ
fcis-6998	44	22	federal	federal	ADJ
fcis-6998	44	23	learning	learning	NOUN
fcis-6998	44	24	and	and	CCONJ
fcis-6998	44	25	vertical	vertical	ADJ
fcis-6998	44	26	federal	federal	ADJ
fcis-6998	44	27	learning	learning	NOUN
fcis-6998	45	1	[	[	X
fcis-6998	45	2	11,12	11,12	NOUN
fcis-6998	45	3	]	]	PUNCT
fcis-6998	45	4	.	.	PUNCT
fcis-6998	46	1	the	the	DET
fcis-6998	46	2	former	former	ADJ
fcis-6998	46	3	mainly	mainly	ADV
fcis-6998	46	4	trains	train	NOUN
fcis-6998	46	5	to	to	PART
fcis-6998	46	6	extract	extract	VERB
fcis-6998	46	7	the	the	DET
fcis-6998	46	8	same	same	ADJ
fcis-6998	46	9	feature	feature	NOUN
fcis-6998	46	10	descriptions	description	NOUN
fcis-6998	46	11	of	of	ADP
fcis-6998	46	12	different	different	ADJ
fcis-6998	46	13	objectives	objective	NOUN
fcis-6998	46	14	from	from	ADP
fcis-6998	46	15	multiple	multiple	ADJ
fcis-6998	46	16	aspects	aspect	NOUN
fcis-6998	46	17	;	;	PUNCT
fcis-6998	46	18	the	the	DET
fcis-6998	46	19	latter	latter	ADJ
fcis-6998	46	20	mainly	mainly	ADV
fcis-6998	46	21	trains	train	NOUN
fcis-6998	46	22	and	and	CCONJ
fcis-6998	46	23	extracts	extract	NOUN
fcis-6998	46	24	the	the	DET
fcis-6998	46	25	same	same	ADJ
fcis-6998	46	26	feature	feature	NOUN
fcis-6998	46	27	description	description	NOUN
fcis-6998	46	28	of	of	ADP
fcis-6998	46	29	the	the	DET
fcis-6998	46	30	same	same	ADJ
fcis-6998	46	31	target	target	NOUN
fcis-6998	46	32	from	from	ADP
fcis-6998	46	33	multiple	multiple	ADJ
fcis-6998	46	34	aspects	aspect	NOUN
fcis-6998	46	35	;	;	PUNCT
fcis-6998	46	36	the	the	DET
fcis-6998	46	37	latter	latter	ADJ
fcis-6998	46	38	mainly	mainly	ADV
fcis-6998	46	39	trains	train	VERB
fcis-6998	46	40	federal	federal	ADJ
fcis-6998	46	41	migration	migration	NOUN
fcis-6998	46	42	learning	learn	VERB
fcis-6998	46	43	to	to	PART
fcis-6998	46	44	solve	solve	VERB
fcis-6998	46	45	the	the	DET
fcis-6998	46	46	problem	problem	NOUN
fcis-6998	46	47	of	of	ADP
fcis-6998	46	48	insufficient	insufficient	ADJ
fcis-6998	46	49	small	small	ADJ
fcis-6998	46	50	label	label	NOUN
fcis-6998	46	51	samples	sample	NOUN
fcis-6998	46	52	and	and	CCONJ
fcis-6998	46	53	data	datum	NOUN
fcis-6998	46	54	sets	set	NOUN
fcis-6998	46	55	.	.	PUNCT
fcis-6998	47	1	as	as	ADP
fcis-6998	47	2	an	an	DET
fcis-6998	47	3	emerging	emerge	VERB
fcis-6998	47	4	technology	technology	NOUN
fcis-6998	47	5	to	to	PART
fcis-6998	47	6	meet	meet	VERB
fcis-6998	47	7	this	this	DET
fcis-6998	47	8	challenge	challenge	NOUN
fcis-6998	47	9	,	,	PUNCT
fcis-6998	47	10	federated	federated	ADJ
fcis-6998	47	11	learning	learning	NOUN
fcis-6998	47	12	has	have	VERB
fcis-6998	47	13	broad	broad	ADJ
fcis-6998	47	14	application	application	NOUN
fcis-6998	47	15	prospects	prospect	NOUN
fcis-6998	47	16	in	in	ADP
fcis-6998	47	17	many	many	ADJ
fcis-6998	47	18	fields	field	NOUN
fcis-6998	47	19	.	.	PUNCT
fcis-6998	48	1	for	for	ADP
fcis-6998	48	2	example	example	NOUN
fcis-6998	48	3	,	,	PUNCT
fcis-6998	48	4	edge	edge	NOUN
fcis-6998	48	5	computing	computing	NOUN
fcis-6998	48	6	and	and	CCONJ
fcis-6998	48	7	the	the	DET
fcis-6998	48	8	internet	internet	NOUN
fcis-6998	48	9	of	of	ADP
fcis-6998	48	10	things	thing	NOUN
fcis-6998	48	11	,	,	PUNCT
fcis-6998	48	12	smart	smart	ADJ
fcis-6998	48	13	medicine	medicine	NOUN
fcis-6998	48	14	,	,	PUNCT
fcis-6998	48	15	financial	financial	ADJ
fcis-6998	48	16	risk	risk	NOUN
fcis-6998	48	17	control	control	NOUN
fcis-6998	48	18	,	,	PUNCT
fcis-6998	48	19	smart	smart	ADJ
fcis-6998	48	20	city	city	NOUN
fcis-6998	48	21	,	,	PUNCT
fcis-6998	48	22	security	security	NOUN
fcis-6998	48	23	sharing	sharing	NOUN
fcis-6998	48	24	,	,	PUNCT
fcis-6998	48	25	etc	etc	X
fcis-6998	48	26	.	.	X
fcis-6998	48	27	in	in	ADP
fcis-6998	48	28	the	the	DET
fcis-6998	48	29	field	field	NOUN
fcis-6998	48	30	of	of	ADP
fcis-6998	48	31	on	on	ADP
fcis-6998	48	32	-	-	PUNCT
fcis-6998	48	33	board	board	NOUN
fcis-6998	48	34	system	system	NOUN
fcis-6998	48	35	intrusion	intrusion	NOUN
fcis-6998	48	36	detection	detection	NOUN
fcis-6998	48	37	systems	system	NOUN
fcis-6998	48	38	,	,	PUNCT
fcis-6998	48	39	[	[	X
fcis-6998	48	40	13	13	NUM
fcis-6998	48	41	]	]	SYM
fcis-6998	48	42	yunlong	yunlong	PROPN
fcis-6998	48	43	lu	lu	PROPN
fcis-6998	48	44	proposed	propose	VERB
fcis-6998	48	45	a	a	DET
fcis-6998	48	46	differential	differential	ADJ
fcis-6998	48	47	private	private	ADJ
fcis-6998	48	48	asynchronous	asynchronous	ADJ
fcis-6998	48	49	federated	federated	ADJ
fcis-6998	48	50	learning	learning	NOUN
fcis-6998	48	51	(	(	PUNCT
fcis-6998	48	52	dpafl	dpafl	NOUN
fcis-6998	48	53	)	)	PUNCT
fcis-6998	48	54	scheme	scheme	NOUN
fcis-6998	48	55	for	for	ADP
fcis-6998	48	56	the	the	DET
fcis-6998	48	57	sharing	sharing	NOUN
fcis-6998	48	58	of	of	ADP
fcis-6998	48	59	automotive	automotive	ADJ
fcis-6998	48	60	network	network	NOUN
fcis-6998	48	61	resources	resource	NOUN
fcis-6998	48	62	.	.	PUNCT
fcis-6998	49	1	in	in	ADP
fcis-6998	49	2	order	order	NOUN
fcis-6998	49	3	to	to	PART
fcis-6998	49	4	establish	establish	VERB
fcis-6998	49	5	a	a	DET
fcis-6998	49	6	secure	secure	ADJ
fcis-6998	49	7	and	and	CCONJ
fcis-6998	49	8	powerful	powerful	ADJ
fcis-6998	49	9	federated	federated	ADJ
fcis-6998	49	10	learning	learning	NOUN
fcis-6998	49	11	scheme	scheme	NOUN
fcis-6998	49	12	,	,	PUNCT
fcis-6998	49	13	the	the	DET
fcis-6998	49	14	privacy	privacy	NOUN
fcis-6998	49	15	of	of	ADP
fcis-6998	49	16	the	the	DET
fcis-6998	49	17	updated	update	VERB
fcis-6998	49	18	local	local	ADJ
fcis-6998	49	19	model	model	NOUN
fcis-6998	49	20	is	be	AUX
fcis-6998	49	21	protected	protect	VERB
fcis-6998	49	22	by	by	ADP
fcis-6998	49	23	incorporating	incorporate	VERB
fcis-6998	49	24	local	local	ADJ
fcis-6998	49	25	differential	differential	NOUN
fcis-6998	49	26	privacy	privacy	NOUN
fcis-6998	49	27	into	into	ADP
fcis-6998	49	28	the	the	DET
fcis-6998	49	29	gradient	gradient	ADJ
fcis-6998	49	30	descent	descent	NOUN
fcis-6998	49	31	training	training	NOUN
fcis-6998	49	32	scheme	scheme	NOUN
fcis-6998	49	33	.	.	PUNCT
fcis-6998	50	1	[	[	X
fcis-6998	50	2	14	14	NUM
fcis-6998	50	3	]	]	PUNCT
fcis-6998	50	4	subsequently	subsequently	ADV
fcis-6998	50	5	,	,	PUNCT
fcis-6998	50	6	yunlonglu	yunlonglu	PROPN
fcis-6998	50	7	proposed	propose	VERB
fcis-6998	50	8	a	a	DET
fcis-6998	50	9	security	security	NOUN
fcis-6998	50	10	scheme	scheme	NOUN
fcis-6998	50	11	based	base	VERB
fcis-6998	50	12	on	on	ADP
fcis-6998	50	13	federated	federated	ADJ
fcis-6998	50	14	learning	learning	NOUN
fcis-6998	50	15	to	to	PART
fcis-6998	50	16	reduce	reduce	VERB
fcis-6998	50	17	data	data	NOUN
fcis-6998	50	18	leakage	leakage	NOUN
fcis-6998	50	19	in	in	ADP
fcis-6998	50	20	vcp	vcp	NOUN
fcis-6998	50	21	.	.	PUNCT
fcis-6998	51	1	in	in	ADP
fcis-6998	51	2	order	order	NOUN
fcis-6998	51	3	to	to	PART
fcis-6998	51	4	protect	protect	VERB
fcis-6998	51	5	the	the	DET
fcis-6998	51	6	privacy	privacy	NOUN
fcis-6998	51	7	of	of	ADP
fcis-6998	51	8	learning	learning	NOUN
fcis-6998	51	9	process	process	NOUN
fcis-6998	51	10	,	,	PUNCT
fcis-6998	51	11	a	a	DET
fcis-6998	51	12	new	new	ADJ
fcis-6998	51	13	random	random	ADJ
fcis-6998	51	14	sub	sub	NOUN
fcis-6998	51	15	rumor	rumor	NOUN
fcis-6998	51	16	update	update	NOUN
fcis-6998	51	17	scheme	scheme	NOUN
fcis-6998	51	18	is	be	AUX
fcis-6998	51	19	proposed	propose	VERB
fcis-6998	51	20	.	.	PUNCT
fcis-6998	52	1	based	base	VERB
fcis-6998	52	2	on	on	ADP
fcis-6998	52	3	professional	professional	ADJ
fcis-6998	52	4	federal	federal	ADJ
fcis-6998	52	5	learning	learning	NOUN
fcis-6998	52	6	,	,	PUNCT
fcis-6998	52	7	a	a	DET
fcis-6998	52	8	two	two	NUM
fcis-6998	52	9	-	-	PUNCT
fcis-6998	52	10	phase	phase	NOUN
fcis-6998	52	11	solution	solution	NOUN
fcis-6998	52	12	is	be	AUX
fcis-6998	52	13	designed	design	VERB
fcis-6998	52	14	to	to	PART
fcis-6998	52	15	reduce	reduce	VERB
fcis-6998	52	16	vehicle	vehicle	NOUN
fcis-6998	52	17	data	datum	NOUN
fcis-6998	52	18	leakage	leakage	NOUN
fcis-6998	52	19	:	:	PUNCT
fcis-6998	52	20	intelligent	intelligent	ADJ
fcis-6998	52	21	data	datum	NOUN
fcis-6998	52	22	conversion	conversion	NOUN
fcis-6998	52	23	and	and	CCONJ
fcis-6998	52	24	collaborative	collaborative	ADJ
fcis-6998	52	25	data	datum	NOUN
fcis-6998	52	26	leakage	leakage	NOUN
fcis-6998	52	27	detection	detection	NOUN
fcis-6998	52	28	.	.	PUNCT
fcis-6998	53	1	this	this	DET
fcis-6998	53	2	scheme	scheme	NOUN
fcis-6998	53	3	can	can	AUX
fcis-6998	53	4	be	be	AUX
fcis-6998	53	5	deployed	deploy	VERB
fcis-6998	53	6	in	in	ADP
fcis-6998	53	7	multi	multi	ADJ
fcis-6998	53	8	-	-	NOUN
fcis-6998	53	9	user	user	NOUN
fcis-6998	53	10	and	and	CCONJ
fcis-6998	53	11	multi	multi	ADJ
fcis-6998	53	12	transmission	transmission	NOUN
fcis-6998	53	13	channel	channel	NOUN
fcis-6998	53	14	application	application	NOUN
fcis-6998	53	15	scenarios	scenario	NOUN
fcis-6998	53	16	.	.	PUNCT
fcis-6998	54	1	3	3	X
fcis-6998	54	2	.	.	X
fcis-6998	54	3	internet	internet	NOUN
fcis-6998	54	4	of	of	ADP
fcis-6998	54	5	vehicles	vehicle	NOUN
fcis-6998	54	6	framework	framework	NOUN
fcis-6998	54	7	based	base	VERB
fcis-6998	54	8	on	on	ADP
fcis-6998	54	9	mobile	mobile	ADJ
fcis-6998	54	10	edge	edge	NOUN
fcis-6998	54	11	computing	computing	NOUN
fcis-6998	54	12	and	and	CCONJ
fcis-6998	54	13	federated	federated	ADJ
fcis-6998	54	14	learning	learn	VERB
fcis-6998	54	15	3.1	3.1	NUM
fcis-6998	54	16	.	.	PUNCT
fcis-6998	55	1	framework	framework	NOUN
fcis-6998	55	2	overview	overview	NOUN
fcis-6998	55	3	in	in	ADP
fcis-6998	55	4	the	the	DET
fcis-6998	55	5	internet	internet	NOUN
fcis-6998	55	6	of	of	ADP
fcis-6998	55	7	vehicles	vehicle	NOUN
fcis-6998	55	8	framework	framework	NOUN
fcis-6998	55	9	based	base	VERB
fcis-6998	55	10	on	on	ADP
fcis-6998	55	11	hierarchical	hierarchical	ADJ
fcis-6998	55	12	federated	federated	ADJ
fcis-6998	55	13	learning	learning	NOUN
fcis-6998	56	1	,	,	PUNCT
fcis-6998	56	2	ecs	ecs	PROPN
fcis-6998	56	3	combines	combine	VERB
fcis-6998	56	4	different	different	ADJ
fcis-6998	56	5	mec	mec	PROPN
fcis-6998	56	6	platforms	platform	NOUN
fcis-6998	56	7	to	to	PART
fcis-6998	56	8	obtain	obtain	VERB
fcis-6998	56	9	traffic	traffic	NOUN
fcis-6998	56	10	data	datum	NOUN
fcis-6998	56	11	from	from	ADP
fcis-6998	56	12	the	the	DET
fcis-6998	56	13	automotive	automotive	ADJ
fcis-6998	56	14	edge	edge	NOUN
fcis-6998	56	15	network	network	NOUN
fcis-6998	56	16	through	through	ADP
fcis-6998	56	17	5	5	NUM
fcis-6998	56	18	g.	g.	PROPN
fcis-6998	56	19	mec	mec	PROPN
fcis-6998	56	20	platform	platform	NOUN
fcis-6998	56	21	is	be	AUX
fcis-6998	56	22	represented	represent	VERB
fcis-6998	56	23	by｛m1	by｛m1	PROPN
fcis-6998	56	24	,	,	PUNCT
fcis-6998	56	25	m2	m2	PROPN
fcis-6998	56	26	,	,	PUNCT
fcis-6998	56	27	...	...	PUNCT
fcis-6998	56	28	,	,	PUNCT
fcis-6998	56	29	mn｝	mn｝	NOUN
fcis-6998	56	30	,	,	PUNCT
fcis-6998	56	31	and	and	CCONJ
fcis-6998	56	32	data	datum	NOUN
fcis-6998	56	33	acquired	acquire	VERB
fcis-6998	56	34	by	by	ADP
fcis-6998	56	35	the	the	DET
fcis-6998	56	36	edge	edge	NOUN
fcis-6998	56	37	automobile	automobile	NOUN
fcis-6998	56	38	network	network	NOUN
fcis-6998	56	39	is	be	AUX
fcis-6998	56	40	represented	represent	VERB
fcis-6998	56	41	by	by	ADP
fcis-6998	56	42	｛d1	｛d1	PROPN
fcis-6998	56	43	,	,	PUNCT
fcis-6998	56	44	d2	d2	PROPN
fcis-6998	56	45	,	,	PUNCT
fcis-6998	56	46	...	...	PUNCT
fcis-6998	56	47	,	,	PUNCT
fcis-6998	56	48	dn｝.	dn｝.	SCONJ
fcis-6998	56	49	based	base	VERB
fcis-6998	56	50	on	on	ADP
fcis-6998	56	51	the	the	DET
fcis-6998	56	52	consideration	consideration	NOUN
fcis-6998	56	53	of	of	ADP
fcis-6998	56	54	data	datum	NOUN
fcis-6998	56	55	privacy	privacy	NOUN
fcis-6998	56	56	,	,	PUNCT
fcis-6998	56	57	we	we	PRON
fcis-6998	56	58	hope	hope	VERB
fcis-6998	56	59	that	that	SCONJ
fcis-6998	56	60	ecs	ecs	PROPN
fcis-6998	56	61	can	can	AUX
fcis-6998	56	62	not	not	PART
fcis-6998	56	63	directly	directly	ADV
fcis-6998	56	64	aggregate	aggregate	VERB
fcis-6998	56	65	all	all	DET
fcis-6998	56	66	data	datum	NOUN
fcis-6998	56	67	d	d	NOUN
fcis-6998	56	68	=	=	SYM
fcis-6998	56	69	d1∪d2∪	d1∪d2∪	NUM
fcis-6998	56	70	...	...	PUNCT
fcis-6998	56	71	∪dn	∪dn	NOUN
fcis-6998	56	72	.	.	PUNCT
fcis-6998	57	1	the	the	DET
fcis-6998	57	2	internet	internet	NOUN
fcis-6998	57	3	of	of	ADP
fcis-6998	57	4	vehicles	vehicle	NOUN
fcis-6998	57	5	framework	framework	NOUN
fcis-6998	57	6	of	of	ADP
fcis-6998	57	7	hierarchical	hierarchical	ADJ
fcis-6998	57	8	federated	federated	ADJ
fcis-6998	57	9	learning	learning	NOUN
fcis-6998	57	10	aims	aim	VERB
fcis-6998	57	11	to	to	PART
fcis-6998	57	12	achieve	achieve	VERB
fcis-6998	57	13	fast	fast	ADJ
fcis-6998	57	14	and	and	CCONJ
fcis-6998	57	15	efficient	efficient	ADJ
fcis-6998	57	16	data	datum	NOUN
fcis-6998	57	17	analysis	analysis	NOUN
fcis-6998	57	18	and	and	CCONJ
fcis-6998	57	19	mining	mine	VERB
fcis-6998	57	20	through	through	ADP
fcis-6998	57	21	federated	federated	ADJ
fcis-6998	57	22	learning	learning	NOUN
fcis-6998	57	23	and	and	CCONJ
fcis-6998	57	24	mec	mec	PROPN
fcis-6998	57	25	platform	platform	NOUN
fcis-6998	57	26	under	under	ADP
fcis-6998	57	27	the	the	DET
fcis-6998	57	28	condition	condition	NOUN
fcis-6998	57	29	of	of	ADP
fcis-6998	57	30	privacy	privacy	NOUN
fcis-6998	57	31	protection	protection	NOUN
fcis-6998	57	32	.	.	PUNCT
fcis-6998	58	1	its	its	PRON
fcis-6998	58	2	frame	frame	NOUN
fcis-6998	58	3	has	have	VERB
fcis-6998	58	4	three	three	NUM
fcis-6998	58	5	layers	layer	NOUN
fcis-6998	58	6	,	,	PUNCT
fcis-6998	58	7	as	as	SCONJ
fcis-6998	58	8	shown	show	VERB
fcis-6998	58	9	in	in	ADP
fcis-6998	58	10	fig	fig	NOUN
fcis-6998	58	11	1	1	NUM
fcis-6998	58	12	.	.	PUNCT
fcis-6998	59	1	the	the	DET
fcis-6998	59	2	top	top	ADJ
fcis-6998	59	3	layer	layer	NOUN
fcis-6998	59	4	is	be	AUX
fcis-6998	59	5	the	the	DET
fcis-6998	59	6	ecs	ecs	PROPN
fcis-6998	59	7	of	of	ADP
fcis-6998	59	8	5	5	NUM
fcis-6998	59	9	g	g	PROPN
fcis-6998	59	10	operators	operator	NOUN
fcis-6998	59	11	,	,	PUNCT
fcis-6998	59	12	which	which	PRON
fcis-6998	59	13	has	have	VERB
fcis-6998	59	14	a	a	DET
fcis-6998	59	15	large	large	ADJ
fcis-6998	59	16	amount	amount	NOUN
fcis-6998	59	17	of	of	ADP
fcis-6998	59	18	data	data	NOUN
fcis-6998	59	19	computing	computing	NOUN
fcis-6998	59	20	and	and	CCONJ
fcis-6998	59	21	processing	processing	NOUN
fcis-6998	59	22	capabilities	capability	NOUN
fcis-6998	59	23	for	for	ADP
fcis-6998	59	24	training	train	VERB
fcis-6998	59	25	the	the	DET
fcis-6998	59	26	server	server	NOUN
fcis-6998	59	27	model	model	NOUN
fcis-6998	59	28	.	.	PUNCT
fcis-6998	60	1	the	the	DET
fcis-6998	60	2	bottom	bottom	ADJ
fcis-6998	60	3	layer	layer	NOUN
fcis-6998	60	4	is	be	AUX
fcis-6998	60	5	the	the	DET
fcis-6998	60	6	edge	edge	NOUN
fcis-6998	60	7	car	car	NOUN
fcis-6998	60	8	node	node	NOUN
fcis-6998	60	9	,	,	PUNCT
fcis-6998	60	10	which	which	PRON
fcis-6998	60	11	is	be	AUX
fcis-6998	60	12	widely	widely	ADV
fcis-6998	60	13	distributed	distribute	VERB
fcis-6998	60	14	and	and	CCONJ
fcis-6998	60	15	communicates	communicate	VERB
fcis-6998	60	16	with	with	ADP
fcis-6998	60	17	each	each	DET
fcis-6998	60	18	other	other	ADJ
fcis-6998	60	19	.	.	PUNCT
fcis-6998	61	1	the	the	DET
fcis-6998	61	2	mec	mec	PROPN
fcis-6998	61	3	platform	platform	NOUN
fcis-6998	61	4	of	of	ADP
fcis-6998	61	5	the	the	DET
fcis-6998	61	6	middle	middle	ADJ
fcis-6998	61	7	layer	layer	NOUN
fcis-6998	61	8	can	can	AUX
fcis-6998	61	9	obtain	obtain	VERB
fcis-6998	61	10	model	model	NOUN
fcis-6998	61	11	parameters	parameter	NOUN
fcis-6998	61	12	from	from	ADP
fcis-6998	61	13	the	the	DET
fcis-6998	61	14	edge	edge	NOUN
fcis-6998	61	15	car	car	NOUN
fcis-6998	61	16	nodes	node	NOUN
fcis-6998	61	17	to	to	PART
fcis-6998	61	18	aggregate	aggregate	VERB
fcis-6998	61	19	models	model	NOUN
fcis-6998	61	20	.	.	PUNCT
fcis-6998	62	1	mec	mec	PROPN
fcis-6998	62	2	is	be	AUX
fcis-6998	62	3	not	not	PART
fcis-6998	62	4	limited	limit	VERB
fcis-6998	62	5	by	by	ADP
fcis-6998	62	6	resources	resource	NOUN
fcis-6998	62	7	and	and	CCONJ
fcis-6998	62	8	computing	computing	NOUN
fcis-6998	62	9	capacity	capacity	NOUN
fcis-6998	62	10	,	,	PUNCT
fcis-6998	62	11	so	so	CCONJ
fcis-6998	62	12	the	the	DET
fcis-6998	62	13	internet	internet	NOUN
fcis-6998	62	14	of	of	ADP
fcis-6998	62	15	vehicles	vehicle	NOUN
fcis-6998	62	16	framework	framework	NOUN
fcis-6998	62	17	based	base	VERB
fcis-6998	62	18	on	on	ADP
fcis-6998	62	19	hierarchical	hierarchical	ADJ
fcis-6998	62	20	federated	federated	ADJ
fcis-6998	62	21	learning	learning	NOUN
fcis-6998	62	22	should	should	AUX
fcis-6998	62	23	be	be	AUX
fcis-6998	62	24	able	able	ADJ
fcis-6998	62	25	to	to	PART
fcis-6998	62	26	store	store	VERB
fcis-6998	62	27	and	and	CCONJ
fcis-6998	62	28	process	process	NOUN
fcis-6998	62	29	data	datum	NOUN
fcis-6998	62	30	from	from	ADP
fcis-6998	62	31	all	all	DET
fcis-6998	62	32	sensor	sensor	NOUN
fcis-6998	62	33	networks	network	NOUN
fcis-6998	62	34	,	,	PUNCT
fcis-6998	62	35	and	and	CCONJ
fcis-6998	62	36	should	should	AUX
fcis-6998	62	37	provide	provide	VERB
fcis-6998	62	38	rapid	rapid	ADJ
fcis-6998	62	39	response	response	NOUN
fcis-6998	62	40	in	in	ADP
fcis-6998	62	41	a	a	DET
fcis-6998	62	42	short	short	ADJ
fcis-6998	62	43	time	time	NOUN
fcis-6998	62	44	.	.	PUNCT
fcis-6998	63	1	according	accord	VERB
fcis-6998	63	2	to	to	ADP
fcis-6998	63	3	this	this	DET
fcis-6998	63	4	new	new	ADJ
fcis-6998	63	5	system	system	NOUN
fcis-6998	63	6	framework	framework	NOUN
fcis-6998	63	7	,	,	PUNCT
fcis-6998	63	8	using	use	VERB
fcis-6998	63	9	the	the	DET
fcis-6998	63	10	hierfedprox	hierfedprox	ADJ
fcis-6998	63	11	algorithm	algorithm	NOUN
fcis-6998	63	12	,	,	PUNCT
fcis-6998	63	13	the	the	DET
fcis-6998	63	14	key	key	ADJ
fcis-6998	63	15	steps	step	NOUN
fcis-6998	63	16	are	be	AUX
fcis-6998	63	17	as	as	SCONJ
fcis-6998	63	18	follows	follow	VERB
fcis-6998	63	19	:	:	PUNCT
fcis-6998	63	20	τ	τ	PROPN
fcis-6998	63	21	after	after	ADP
fcis-6998	63	22	a	a	DET
fcis-6998	63	23	local	local	ADJ
fcis-6998	63	24	update	update	NOUN
fcis-6998	63	25	,	,	PUNCT
fcis-6998	63	26	each	each	DET
fcis-6998	63	27	edge	edge	NOUN
fcis-6998	63	28	server	server	NOUN
fcis-6998	63	29	aggregates	aggregate	VERB
fcis-6998	63	30	the	the	DET
fcis-6998	63	31	training	training	NOUN
fcis-6998	63	32	models	model	NOUN
fcis-6998	63	33	of	of	ADP
fcis-6998	63	34	all	all	DET
fcis-6998	63	35	its	its	PRON
fcis-6998	63	36	edge	edge	NOUN
fcis-6998	63	37	car	car	NOUN
fcis-6998	63	38	nodes	node	NOUN
fcis-6998	63	39	.	.	PUNCT
fcis-6998	64	1	then	then	ADV
fcis-6998	64	2	at	at	ADP
fcis-6998	64	3	each	each	DET
fcis-6998	64	4	τ	τ	NOUN
fcis-6998	64	5	after	after	ADP
fcis-6998	64	6	two	two	NUM
fcis-6998	64	7	edge	edge	NOUN
fcis-6998	64	8	aggregation	aggregation	NOUN
fcis-6998	64	9	models	model	NOUN
fcis-6998	64	10	,	,	PUNCT
fcis-6998	64	11	ecs	ecs	PROPN
fcis-6998	64	12	aggregates	aggregate	VERB
fcis-6998	64	13	all	all	DET
fcis-6998	64	14	edge	edge	NOUN
fcis-6998	64	15	server	server	NOUN
fcis-6998	64	16	models	model	NOUN
fcis-6998	64	17	.	.	PUNCT
fcis-6998	65	1	the	the	DET
fcis-6998	65	2	data	datum	NOUN
fcis-6998	65	3	and	and	CCONJ
fcis-6998	65	4	information	information	NOUN
fcis-6998	65	5	of	of	ADP
fcis-6998	65	6	any	any	DET
fcis-6998	65	7	edge	edge	NOUN
fcis-6998	65	8	vehicle	vehicle	NOUN
fcis-6998	65	9	node	node	NOUN
fcis-6998	65	10	will	will	AUX
fcis-6998	65	11	not	not	PART
fcis-6998	65	12	be	be	AUX
fcis-6998	65	13	disclosed	disclose	VERB
fcis-6998	65	14	in	in	ADP
fcis-6998	65	15	the	the	DET
fcis-6998	65	16	whole	whole	ADJ
fcis-6998	65	17	process	process	NOUN
fcis-6998	65	18	.	.	PUNCT
fcis-6998	66	1	figure	figure	NOUN
fcis-6998	66	2	1	1	NUM
fcis-6998	66	3	.	.	PUNCT
fcis-6998	66	4	hierarchical	hierarchical	ADJ
fcis-6998	66	5	federal	federal	ADJ
fcis-6998	66	6	learning	learn	VERB
fcis-6998	66	7	architecture	architecture	NOUN
fcis-6998	66	8	of	of	ADP
fcis-6998	66	9	internet	internet	NOUN
fcis-6998	66	10	of	of	ADP
fcis-6998	66	11	vehicles	vehicle	NOUN
fcis-6998	66	12	3.2	3.2	NUM
fcis-6998	66	13	.	.	PUNCT
fcis-6998	67	1	edge	edge	NOUN
fcis-6998	67	2	car	car	NOUN
fcis-6998	67	3	node	node	NOUN
fcis-6998	67	4	edge	edge	NOUN
fcis-6998	67	5	-cloud	-cloud	PROPN
fcis-6998	67	6	three	three	NUM
fcis-6998	67	7	-	-	PUNCT
fcis-6998	67	8	layer	layer	NOUN
fcis-6998	67	9	federated	federated	ADJ
fcis-6998	67	10	learning	learning	NOUN
fcis-6998	67	11	in	in	ADP
fcis-6998	67	12	the	the	DET
fcis-6998	67	13	federated	federated	ADJ
fcis-6998	67	14	learning	learning	NOUN
fcis-6998	67	15	based	base	VERB
fcis-6998	67	16	on	on	ADP
fcis-6998	67	17	cloud	cloud	ADJ
fcis-6998	67	18	edge	edge	NOUN
fcis-6998	67	19	automobile	automobile	NOUN
fcis-6998	67	20	nodes	node	NOUN
fcis-6998	67	21	,	,	PUNCT
fcis-6998	67	22	the	the	DET
fcis-6998	67	23	ecs	ec	NOUN
fcis-6998	67	24	can	can	AUX
fcis-6998	67	25	aggregate	aggregate	VERB
fcis-6998	67	26	the	the	DET
fcis-6998	67	27	data	datum	NOUN
fcis-6998	67	28	of	of	ADP
fcis-6998	67	29	many	many	ADJ
fcis-6998	67	30	clients	client	NOUN
fcis-6998	67	31	,	,	PUNCT
fcis-6998	67	32	but	but	CCONJ
fcis-6998	67	33	the	the	DET
fcis-6998	67	34	communication	communication	NOUN
fcis-6998	67	35	cost	cost	NOUN
fcis-6998	67	36	is	be	AUX
fcis-6998	67	37	large	large	ADJ
fcis-6998	67	38	.	.	PUNCT
fcis-6998	68	1	the	the	DET
fcis-6998	68	2	edge	edge	NOUN
fcis-6998	68	3	server	server	NOUN
fcis-6998	68	4	can	can	AUX
fcis-6998	68	5	only	only	ADV
fcis-6998	68	6	aggregate	aggregate	VERB
fcis-6998	68	7	a	a	DET
fcis-6998	68	8	relatively	relatively	ADV
fcis-6998	68	9	small	small	ADJ
fcis-6998	68	10	amount	amount	NOUN
fcis-6998	68	11	of	of	ADP
fcis-6998	68	12	data	datum	NOUN
fcis-6998	68	13	from	from	ADP
fcis-6998	68	14	edge	edge	NOUN
fcis-6998	68	15	car	car	NOUN
fcis-6998	68	16	nodes	node	NOUN
fcis-6998	68	17	,	,	PUNCT
fcis-6998	68	18	but	but	CCONJ
fcis-6998	68	19	the	the	DET
fcis-6998	68	20	communication	communication	NOUN
fcis-6998	68	21	cost	cost	NOUN
fcis-6998	68	22	is	be	AUX
fcis-6998	68	23	low	low	ADJ
fcis-6998	68	24	.	.	PUNCT
fcis-6998	69	1	combining	combine	VERB
fcis-6998	69	2	the	the	DET
fcis-6998	69	3	advantages	advantage	NOUN
fcis-6998	69	4	of	of	ADP
fcis-6998	69	5	both	both	PRON
fcis-6998	69	6	,	,	PUNCT
fcis-6998	69	7	we	we	PRON
fcis-6998	69	8	consider	consider	VERB
fcis-6998	69	9	a	a	DET
fcis-6998	69	10	hierarchical	hierarchical	ADJ
fcis-6998	69	11	federated	federated	ADJ
fcis-6998	69	12	learning	learning	NOUN
fcis-6998	69	13	of	of	ADP
fcis-6998	69	14	edge	edge	NOUN
fcis-6998	69	15	car	car	NOUN
fcis-6998	69	16	node	node	NOUN
fcis-6998	69	17	edge	edge	NOUN
fcis-6998	69	18	cloud	cloud	NOUN
fcis-6998	69	19	,	,	PUNCT
fcis-6998	69	20	supported	support	VERB
fcis-6998	69	21	by	by	ADP
fcis-6998	69	22	hierfedprox	hierfedprox	ADJ
fcis-6998	69	23	algorithm	algorithm	NOUN
fcis-6998	69	24	.	.	PUNCT
fcis-6998	70	1	this	this	DET
fcis-6998	70	2	algorithm	algorithm	NOUN
fcis-6998	70	3	[	[	X
fcis-6998	70	4	15,16	15,16	NUM
fcis-6998	70	5	]	]	PUNCT
fcis-6998	70	6	can	can	AUX
fcis-6998	70	7	train	train	VERB
fcis-6998	70	8	the	the	DET
fcis-6998	70	9	37	37	NUM
fcis-6998	70	10	model	model	NOUN
fcis-6998	70	11	faster	fast	ADV
fcis-6998	70	12	,	,	PUNCT
fcis-6998	70	13	achieve	achieve	VERB
fcis-6998	70	14	better	well	ADJ
fcis-6998	70	15	communication	communication	NOUN
fcis-6998	70	16	calculation	calculation	NOUN
fcis-6998	70	17	tradeoffs	tradeoff	NOUN
fcis-6998	70	18	,	,	PUNCT
fcis-6998	70	19	and	and	CCONJ
fcis-6998	70	20	improve	improve	VERB
fcis-6998	70	21	the	the	DET
fcis-6998	70	22	accuracy	accuracy	NOUN
fcis-6998	70	23	of	of	ADP
fcis-6998	70	24	the	the	DET
fcis-6998	70	25	model	model	NOUN
fcis-6998	70	26	scene	scene	NOUN
fcis-6998	70	27	.	.	PUNCT
fcis-6998	71	1	the	the	DET
fcis-6998	71	2	process	process	NOUN
fcis-6998	71	3	of	of	ADP
fcis-6998	71	4	model	model	NOUN
fcis-6998	71	5	learning	learning	NOUN
fcis-6998	71	6	is	be	AUX
fcis-6998	71	7	realized	realize	VERB
fcis-6998	71	8	by	by	ADP
fcis-6998	71	9	minimizing	minimize	VERB
fcis-6998	71	10	a	a	DET
fcis-6998	71	11	loss	loss	NOUN
fcis-6998	71	12	function	function	NOUN
fcis-6998	71	13	,	,	PUNCT
fcis-6998	71	14	which	which	PRON
fcis-6998	71	15	is	be	AUX
fcis-6998	71	16	calculated	calculate	VERB
fcis-6998	71	17	according	accord	VERB
fcis-6998	71	18	to	to	ADP
fcis-6998	71	19	the	the	DET
fcis-6998	71	20	data	datum	NOUN
fcis-6998	71	21	on	on	ADP
fcis-6998	71	22	the	the	DET
fcis-6998	71	23	edge	edge	NOUN
fcis-6998	71	24	car	car	NOUN
fcis-6998	71	25	node	node	NOUN
fcis-6998	71	26	.	.	PUNCT
fcis-6998	72	1	the	the	DET
fcis-6998	72	2	purpose	purpose	NOUN
fcis-6998	72	3	is	be	AUX
fcis-6998	72	4	to	to	PART
fcis-6998	72	5	make	make	VERB
fcis-6998	72	6	the	the	DET
fcis-6998	72	7	local	local	ADJ
fcis-6998	72	8	update	update	NOUN
fcis-6998	72	9	not	not	PART
fcis-6998	72	10	too	too	ADV
fcis-6998	72	11	far	far	ADV
fcis-6998	72	12	away	away	ADV
fcis-6998	72	13	from	from	ADP
fcis-6998	72	14	the	the	DET
fcis-6998	72	15	initial	initial	ADJ
fcis-6998	72	16	global	global	ADJ
fcis-6998	72	17	model	model	NOUN
fcis-6998	72	18	,	,	PUNCT
fcis-6998	72	19	and	and	CCONJ
fcis-6998	72	20	reduce	reduce	VERB
fcis-6998	72	21	the	the	DET
fcis-6998	72	22	impact	impact	NOUN
fcis-6998	72	23	of	of	ADP
fcis-6998	72	24	no	no	DET
fcis-6998	72	25	independent	independent	ADJ
fcis-6998	72	26	identically	identically	ADV
fcis-6998	72	27	distributed	distribute	VERB
fcis-6998	72	28	(	(	PUNCT
fcis-6998	72	29	non	non	ADJ
fcis-6998	72	30	iid	iid	NOUN
fcis-6998	72	31	)	)	PUNCT
fcis-6998	72	32	on	on	ADP
fcis-6998	72	33	the	the	DET
fcis-6998	72	34	premise	premise	NOUN
fcis-6998	72	35	of	of	ADP
fcis-6998	72	36	tolerating	tolerate	VERB
fcis-6998	72	37	system	system	NOUN
fcis-6998	72	38	heterogeneity	heterogeneity	NOUN
fcis-6998	72	39	.	.	PUNCT
fcis-6998	73	1	the	the	DET
fcis-6998	73	2	loss	loss	NOUN
fcis-6998	73	3	function	function	NOUN
fcis-6998	73	4	is	be	AUX
fcis-6998	73	5	used	use	VERB
fcis-6998	73	6	to	to	PART
fcis-6998	73	7	evaluate	evaluate	VERB
fcis-6998	73	8	the	the	DET
fcis-6998	73	9	difference	difference	NOUN
fcis-6998	73	10	between	between	ADP
fcis-6998	73	11	the	the	DET
fcis-6998	73	12	predicted	predict	VERB
fcis-6998	73	13	value	value	NOUN
fcis-6998	73	14	of	of	ADP
fcis-6998	73	15	the	the	DET
fcis-6998	73	16	model	model	NOUN
fcis-6998	73	17	and	and	CCONJ
fcis-6998	73	18	the	the	DET
fcis-6998	73	19	real	real	ADJ
fcis-6998	73	20	value	value	NOUN
fcis-6998	73	21	.	.	PUNCT
fcis-6998	74	1	the	the	PRON
fcis-6998	74	2	better	well	ADJ
fcis-6998	74	3	the	the	DET
fcis-6998	74	4	loss	loss	NOUN
fcis-6998	74	5	function	function	NOUN
fcis-6998	74	6	is	be	AUX
fcis-6998	74	7	,	,	PUNCT
fcis-6998	74	8	the	the	PRON
fcis-6998	74	9	better	well	ADJ
fcis-6998	74	10	the	the	DET
fcis-6998	74	11	model	model	NOUN
fcis-6998	74	12	performance	performance	NOUN
fcis-6998	74	13	is	be	AUX
fcis-6998	74	14	generally	generally	ADV
fcis-6998	74	15	.	.	PUNCT
fcis-6998	75	1	2	2	NUM
fcis-6998	75	2	(	(	PUNCT
fcis-6998	75	3	;	;	PUNCT
fcis-6998	75	4	)	)	PUNCT
fcis-6998	75	5	(	(	PUNCT
fcis-6998	75	6	)	)	PUNCT
fcis-6998	75	7	2	2	NUM
fcis-6998	76	1	i	i	PRON
fcis-6998	76	2	ih	ih	VERB
fcis-6998	76	3	f	f	PROPN
fcis-6998	76	4	w	w	PROPN
fcis-6998	76	5			X
fcis-6998	76	6			X
fcis-6998	76	7			X
fcis-6998	76	8			X
fcis-6998	76	9	=	=	PROPN
fcis-6998	76	10	+	+	CCONJ
fcis-6998	76	11	−	−	PROPN
fcis-6998	76	12	(	(	PUNCT
fcis-6998	76	13	1	1	NUM
fcis-6998	76	14	)	)	PUNCT
fcis-6998	76	15	h	h	NOUN
fcis-6998	76	16	(	(	PUNCT
fcis-6998	76	17	;	;	PUNCT
fcis-6998	76	18	)	)	PUNCT
fcis-6998	76	19	(	(	PUNCT
fcis-6998	76	20	)	)	PUNCT
fcis-6998	76	21	i	i	PRON
fcis-6998	76	22	if	if	SCONJ
fcis-6998	76	23	w	w	PROPN
fcis-6998	76	24			PROPN
fcis-6998	76	25			PROPN
fcis-6998	76	26			NOUN
fcis-6998	76	27			X
fcis-6998	76	28			NOUN
fcis-6998	77	1	=	=	NOUN
fcis-6998	77	2			NOUN
fcis-6998	77	3	+	+	CCONJ
fcis-6998	77	4	−	−	PROPN
fcis-6998	77	5	（	（	SYM
fcis-6998	77	6	）	）	PUNCT
fcis-6998	77	7	(	(	PUNCT
fcis-6998	77	8	2	2	NUM
fcis-6998	77	9	)	)	PUNCT
fcis-6998	77	10	among	among	ADP
fcis-6998	77	11	them	they	PRON
fcis-6998	77	12	,	,	PUNCT
fcis-6998	77	13	is	be	AUX
fcis-6998	77	14	the	the	DET
fcis-6998	77	15	loss	loss	NOUN
fcis-6998	77	16	function	function	NOUN
fcis-6998	77	17	of	of	ADP
fcis-6998	77	18	the	the	DET
fcis-6998	77	19	federal	federal	ADJ
fcis-6998	77	20	average	average	ADJ
fcis-6998	77	21	algorithm	algorithm	NOUN
fcis-6998	77	22	,	,	PUNCT
fcis-6998	77	23	is	be	AUX
fcis-6998	77	24	the	the	DET
fcis-6998	77	25	local	local	ADJ
fcis-6998	77	26	model	model	NOUN
fcis-6998	77	27	update	update	NOUN
fcis-6998	77	28	weight	weight	NOUN
fcis-6998	77	29	,	,	PUNCT
fcis-6998	77	30	is	be	AUX
fcis-6998	77	31	the	the	DET
fcis-6998	77	32	global	global	ADJ
fcis-6998	77	33	model	model	NOUN
fcis-6998	77	34	weight	weight	NOUN
fcis-6998	77	35	issued	issue	VERB
fcis-6998	77	36	by	by	ADP
fcis-6998	77	37	ecs	ec	NOUN
fcis-6998	77	38	,	,	PUNCT
fcis-6998	77	39	and	and	CCONJ
fcis-6998	77	40	is	be	AUX
fcis-6998	77	41	the	the	DET
fcis-6998	77	42	near	near	ADJ
fcis-6998	77	43	end	end	NOUN
fcis-6998	77	44	penalty	penalty	NOUN
fcis-6998	77	45	coefficient	coefficient	NOUN
fcis-6998	77	46	.	.	PUNCT
fcis-6998	78	1	when	when	SCONJ
fcis-6998	78	2	τ｜τ	τ｜τ	PROPN
fcis-6998	78	3	1	1	NUM
fcis-6998	78	4	≠	≠	PROPN
fcis-6998	78	5	0	0	NUM
fcis-6998	78	6	;	;	PUNCT
fcis-6998	78	7	the	the	DET
fcis-6998	78	8	evolution	evolution	NOUN
fcis-6998	78	9	of	of	ADP
fcis-6998	78	10	local	local	ADJ
fcis-6998	78	11	model	model	NOUN
fcis-6998	78	12	parameters	parameter	NOUN
fcis-6998	78	13	is	be	AUX
fcis-6998	78	14	as	as	SCONJ
fcis-6998	78	15	follows	follow	VERB
fcis-6998	78	16	:	:	PUNCT
fcis-6998	78	17	h	h	NOUN
fcis-6998	78	18	(	(	PUNCT
fcis-6998	78	19	;	;	PUNCT
fcis-6998	78	20	)	)	PUNCT
fcis-6998	79	1	(	(	PUNCT
fcis-6998	79	2	)	)	PUNCT
fcis-6998	79	3	i	i	PRON
fcis-6998	79	4	if	if	SCONJ
fcis-6998	79	5	w	w	PROPN
fcis-6998	79	6			PROPN
fcis-6998	79	7			PROPN
fcis-6998	79	8			NOUN
fcis-6998	79	9			X
fcis-6998	79	10			NOUN
fcis-6998	80	1	=	=	NOUN
fcis-6998	80	2			NOUN
fcis-6998	80	3	+	+	CCONJ
fcis-6998	80	4	−	−	PROPN
fcis-6998	80	5	（	（	SYM
fcis-6998	80	6	）	）	PUNCT
fcis-6998	80	7	(	(	PUNCT
fcis-6998	80	8	3	3	X
fcis-6998	80	9	)	)	PUNCT
fcis-6998	80	10	4	4	NUM
fcis-6998	80	11	.	.	PUNCT
fcis-6998	80	12	experiment	experiment	NOUN
fcis-6998	80	13	this	this	DET
fcis-6998	80	14	section	section	NOUN
fcis-6998	80	15	will	will	AUX
fcis-6998	80	16	introduce	introduce	VERB
fcis-6998	80	17	the	the	DET
fcis-6998	80	18	experimental	experimental	ADJ
fcis-6998	80	19	environment	environment	NOUN
fcis-6998	80	20	,	,	PUNCT
fcis-6998	80	21	experimental	experimental	ADJ
fcis-6998	80	22	design	design	NOUN
fcis-6998	80	23	,	,	PUNCT
fcis-6998	80	24	relevant	relevant	ADJ
fcis-6998	80	25	models	model	NOUN
fcis-6998	80	26	and	and	CCONJ
fcis-6998	80	27	results	result	NOUN
fcis-6998	80	28	.	.	PUNCT
fcis-6998	81	1	adjust	adjust	VERB
fcis-6998	81	2	the	the	DET
fcis-6998	81	3	parameters	parameter	NOUN
fcis-6998	81	4	of	of	ADP
fcis-6998	81	5	hierfedprox	hierfedprox	PROPN
fcis-6998	81	6	,	,	PUNCT
fcis-6998	81	7	and	and	CCONJ
fcis-6998	81	8	make	make	VERB
fcis-6998	81	9	an	an	DET
fcis-6998	81	10	overall	overall	ADJ
fcis-6998	81	11	comparison	comparison	NOUN
fcis-6998	81	12	between	between	ADP
fcis-6998	81	13	hierfedprox	hierfedprox	PROPN
fcis-6998	81	14	and	and	CCONJ
fcis-6998	81	15	fedprox	fedprox	NOUN
fcis-6998	81	16	.	.	PUNCT
fcis-6998	82	1	4.1	4.1	NUM
fcis-6998	82	2	.	.	PUNCT
fcis-6998	82	3	experimental	experimental	ADJ
fcis-6998	82	4	environment	environment	NOUN
fcis-6998	82	5	and	and	CCONJ
fcis-6998	82	6	design	design	VERB
fcis-6998	82	7	the	the	DET
fcis-6998	82	8	experimental	experimental	ADJ
fcis-6998	82	9	computer	computer	NOUN
fcis-6998	82	10	is	be	AUX
fcis-6998	82	11	configured	configure	VERB
fcis-6998	82	12	with	with	ADP
fcis-6998	82	13	16	16	NUM
fcis-6998	82	14	gb	gb	NOUN
fcis-6998	82	15	ram	ram	NOUN
fcis-6998	82	16	,	,	PUNCT
fcis-6998	82	17	i5	i5	NOUN
fcis-6998	82	18	-	-	PUNCT
fcis-6998	82	19	10400cpu@2.90ghz	10400cpu@2.90ghz	NUM
fcis-6998	82	20	the	the	DET
fcis-6998	82	21	processor	processor	NOUN
fcis-6998	82	22	,	,	PUNCT
fcis-6998	82	23	operating	operating	NOUN
fcis-6998	82	24	system	system	NOUN
fcis-6998	82	25	is	be	AUX
fcis-6998	82	26	window	window	NOUN
fcis-6998	82	27	10	10	NUM
fcis-6998	82	28	,	,	PUNCT
fcis-6998	82	29	and	and	CCONJ
fcis-6998	82	30	the	the	DET
fcis-6998	82	31	experimental	experimental	ADJ
fcis-6998	82	32	code	code	NOUN
fcis-6998	82	33	is	be	AUX
fcis-6998	82	34	implemented	implement	VERB
fcis-6998	82	35	using	use	VERB
fcis-6998	82	36	the	the	DET
fcis-6998	82	37	python	python	NOUN
fcis-6998	82	38	framework	framework	NOUN
fcis-6998	82	39	.	.	PUNCT
fcis-6998	83	1	we	we	PRON
fcis-6998	83	2	have	have	AUX
fcis-6998	83	3	conceived	conceive	VERB
fcis-6998	83	4	a	a	DET
fcis-6998	83	5	three	three	NUM
fcis-6998	83	6	-	-	PUNCT
fcis-6998	83	7	layer	layer	NOUN
fcis-6998	83	8	federated	federate	VERB
fcis-6998	83	9	learning	learn	VERB
fcis-6998	83	10	system	system	NOUN
fcis-6998	83	11	of	of	ADP
fcis-6998	83	12	edge	edge	NOUN
fcis-6998	83	13	car	car	NOUN
fcis-6998	83	14	network	network	NOUN
fcis-6998	83	15	edge	edge	NOUN
fcis-6998	83	16	server	server	PROPN
fcis-6998	83	17	cloud	cloud	PROPN
fcis-6998	83	18	server	server	NOUN
fcis-6998	83	19	.	.	PUNCT
fcis-6998	84	1	the	the	DET
fcis-6998	84	2	system	system	NOUN
fcis-6998	84	3	has	have	VERB
fcis-6998	84	4	a	a	DET
fcis-6998	84	5	total	total	NOUN
fcis-6998	84	6	of	of	ADP
fcis-6998	84	7	50	50	NUM
fcis-6998	84	8	edge	edge	NOUN
fcis-6998	84	9	car	car	NOUN
fcis-6998	84	10	nodes	node	NOUN
fcis-6998	84	11	,	,	PUNCT
fcis-6998	84	12	5	5	NUM
fcis-6998	84	13	mec	mec	PROPN
fcis-6998	84	14	edge	edge	NOUN
fcis-6998	84	15	servers	server	NOUN
fcis-6998	84	16	,	,	PUNCT
fcis-6998	84	17	and	and	CCONJ
fcis-6998	84	18	1	1	NUM
fcis-6998	84	19	ecs	ec	NOUN
fcis-6998	84	20	.	.	PUNCT
fcis-6998	84	21	and	and	CCONJ
fcis-6998	84	22	suppose	suppose	VERB
fcis-6998	84	23	that	that	SCONJ
fcis-6998	84	24	each	each	DET
fcis-6998	84	25	mec	mec	PROPN
fcis-6998	84	26	edge	edge	NOUN
fcis-6998	84	27	server	server	NOUN
fcis-6998	84	28	has	have	VERB
fcis-6998	84	29	the	the	DET
fcis-6998	84	30	same	same	ADJ
fcis-6998	84	31	number	number	NOUN
fcis-6998	84	32	of	of	ADP
fcis-6998	84	33	edge	edge	NOUN
fcis-6998	84	34	car	car	NOUN
fcis-6998	84	35	nodes	node	NOUN
fcis-6998	84	36	and	and	CCONJ
fcis-6998	84	37	the	the	DET
fcis-6998	84	38	same	same	ADJ
fcis-6998	84	39	training	training	NOUN
fcis-6998	84	40	data	datum	NOUN
fcis-6998	84	41	.	.	PUNCT
fcis-6998	85	1	based	base	VERB
fcis-6998	85	2	on	on	ADP
fcis-6998	85	3	the	the	DET
fcis-6998	85	4	open	open	ADJ
fcis-6998	85	5	source	source	NOUN
fcis-6998	85	6	mnist	mnist	NOUN
fcis-6998	85	7	dataset	dataset	NOUN
fcis-6998	85	8	,	,	PUNCT
fcis-6998	85	9	local	local	ADJ
fcis-6998	85	10	training	training	NOUN
fcis-6998	85	11	is	be	AUX
fcis-6998	85	12	conducted	conduct	VERB
fcis-6998	85	13	at	at	ADP
fcis-6998	85	14	each	each	DET
fcis-6998	85	15	edge	edge	NOUN
fcis-6998	85	16	vehicle	vehicle	NOUN
fcis-6998	85	17	node	node	NOUN
fcis-6998	85	18	.	.	PUNCT
fcis-6998	86	1	the	the	DET
fcis-6998	86	2	batch	batch	NOUN
fcis-6998	86	3	size	size	NOUN
fcis-6998	86	4	is	be	AUX
fcis-6998	86	5	20	20	NUM
fcis-6998	86	6	,	,	PUNCT
fcis-6998	86	7	the	the	DET
fcis-6998	86	8	initial	initial	ADJ
fcis-6998	86	9	learning	learning	NOUN
fcis-6998	86	10	rate	rate	NOUN
fcis-6998	86	11	is	be	AUX
fcis-6998	86	12	0	0	NUM
fcis-6998	86	13	,	,	PUNCT
fcis-6998	86	14	and	and	CCONJ
fcis-6998	86	15	the	the	DET
fcis-6998	86	16	exponential	exponential	ADJ
fcis-6998	86	17	decay	decay	NOUN
fcis-6998	86	18	rate	rate	NOUN
fcis-6998	86	19	of	of	ADP
fcis-6998	86	20	each	each	DET
fcis-6998	86	21	iteration	iteration	NOUN
fcis-6998	86	22	is	be	AUX
fcis-6998	86	23	0.995	0.995	NUM
fcis-6998	86	24	.	.	PUNCT
fcis-6998	87	1	the	the	DET
fcis-6998	87	2	experimental	experimental	ADJ
fcis-6998	87	3	results	result	NOUN
fcis-6998	87	4	verify	verify	VERB
fcis-6998	87	5	the	the	DET
fcis-6998	87	6	effectiveness	effectiveness	NOUN
fcis-6998	87	7	of	of	ADP
fcis-6998	87	8	this	this	DET
fcis-6998	87	9	method	method	NOUN
fcis-6998	87	10	.	.	PUNCT
fcis-6998	88	1	in	in	ADP
fcis-6998	88	2	the	the	DET
fcis-6998	88	3	training	training	NOUN
fcis-6998	88	4	process	process	NOUN
fcis-6998	88	5	of	of	ADP
fcis-6998	88	6	hierfedprox	hierfedprox	NOUN
fcis-6998	88	7	,	,	PUNCT
fcis-6998	88	8	and	and	CCONJ
fcis-6998	88	9	will	will	AUX
fcis-6998	88	10	not	not	PART
fcis-6998	88	11	only	only	ADV
fcis-6998	88	12	affect	affect	VERB
fcis-6998	88	13	the	the	DET
fcis-6998	88	14	training	training	NOUN
fcis-6998	88	15	accuracy	accuracy	NOUN
fcis-6998	88	16	,	,	PUNCT
fcis-6998	88	17	but	but	CCONJ
fcis-6998	88	18	also	also	ADV
fcis-6998	88	19	directly	directly	ADV
fcis-6998	88	20	affect	affect	VERB
fcis-6998	88	21	the	the	DET
fcis-6998	88	22	computing	computing	NOUN
fcis-6998	88	23	cost	cost	NOUN
fcis-6998	88	24	and	and	CCONJ
fcis-6998	88	25	communication	communication	NOUN
fcis-6998	88	26	efficiency	efficiency	NOUN
fcis-6998	88	27	of	of	ADP
fcis-6998	88	28	edge	edge	NOUN
fcis-6998	88	29	car	car	NOUN
fcis-6998	88	30	nodes	node	NOUN
fcis-6998	88	31	.	.	PUNCT
fcis-6998	89	1	different	different	ADJ
fcis-6998	89	2	sums	sum	NOUN
fcis-6998	89	3	are	be	AUX
fcis-6998	89	4	used	use	VERB
fcis-6998	89	5	in	in	ADP
fcis-6998	89	6	hierfedprox	hierfedprox	PROPN
fcis-6998	89	7	to	to	PART
fcis-6998	89	8	evaluate	evaluate	VERB
fcis-6998	89	9	performance	performance	NOUN
fcis-6998	89	10	.	.	PUNCT
fcis-6998	90	1	4.2	4.2	NUM
fcis-6998	90	2	.	.	PUNCT
fcis-6998	91	1	experimental	experimental	ADJ
fcis-6998	91	2	results	result	NOUN
fcis-6998	91	3	and	and	CCONJ
fcis-6998	91	4	analysis	analysis	NOUN
fcis-6998	91	5	the	the	DET
fcis-6998	91	6	experiment	experiment	NOUN
fcis-6998	91	7	is	be	AUX
fcis-6998	91	8	nothing	nothing	PRON
fcis-6998	91	9	more	more	ADJ
fcis-6998	91	10	than	than	ADP
fcis-6998	91	11	machine	machine	NOUN
fcis-6998	91	12	learning	learn	VERB
fcis-6998	91	13	mnist	mnist	NOUN
fcis-6998	91	14	datasets	dataset	NOUN
fcis-6998	91	15	in	in	ADP
fcis-6998	91	16	two	two	NUM
fcis-6998	91	17	non	non	ADJ
fcis-6998	91	18	iid	iid	VERB
fcis-6998	91	19	scenarios	scenario	NOUN
fcis-6998	91	20	,	,	PUNCT
fcis-6998	91	21	namely	namely	ADV
fcis-6998	91	22	edge	edge	VERB
fcis-6998	91	23	independent	independent	ADJ
fcis-6998	91	24	i	i	PROPN
fcis-6998	91	25	d	d	PROPN
fcis-6998	91	26	and	and	CCONJ
fcis-6998	91	27	edge	edge	VERB
fcis-6998	91	28	non	non	ADJ
fcis-6998	91	29	iid	iid	NOUN
fcis-6998	91	30	.	.	PUNCT
fcis-6998	92	1	first	first	ADV
fcis-6998	92	2	,	,	PUNCT
fcis-6998	92	3	when	when	SCONJ
fcis-6998	92	4	the	the	DET
fcis-6998	92	5	communication	communication	NOUN
fcis-6998	92	6	frequency	frequency	NOUN
fcis-6998	92	7	between	between	ADP
fcis-6998	92	8	the	the	DET
fcis-6998	92	9	mobile	mobile	ADJ
fcis-6998	92	10	vehicle	vehicle	NOUN
fcis-6998	92	11	client	client	NOUN
fcis-6998	92	12	and	and	CCONJ
fcis-6998	92	13	the	the	DET
fcis-6998	92	14	ecs	ecs	PROPN
fcis-6998	92	15	is	be	AUX
fcis-6998	92	16	fixed	fix	VERB
fcis-6998	92	17	,	,	PUNCT
fcis-6998	92	18	that	that	ADV
fcis-6998	92	19	is	is	ADV
fcis-6998	92	20	,	,	PUNCT
fcis-6998	92	21	fixed	fixed	ADJ
fcis-6998	92	22	time	time	NOUN
fcis-6998	92	23	,	,	PUNCT
fcis-6998	92	24	the	the	DET
fcis-6998	92	25	more	more	ADV
fcis-6998	92	26	frequent	frequent	ADJ
fcis-6998	92	27	the	the	DET
fcis-6998	92	28	communication	communication	NOUN
fcis-6998	92	29	between	between	ADP
fcis-6998	92	30	the	the	DET
fcis-6998	92	31	mobile	mobile	ADJ
fcis-6998	92	32	vehicle	vehicle	NOUN
fcis-6998	92	33	client	client	NOUN
fcis-6998	92	34	and	and	CCONJ
fcis-6998	92	35	the	the	DET
fcis-6998	92	36	edge	edge	NOUN
fcis-6998	92	37	server	server	PROPN
fcis-6998	92	38	mec	mec	PROPN
fcis-6998	92	39	is	be	AUX
fcis-6998	92	40	,	,	PUNCT
fcis-6998	92	41	that	that	ADV
fcis-6998	92	42	is	is	ADV
fcis-6998	92	43	,	,	PUNCT
fcis-6998	92	44	reduced	reduce	VERB
fcis-6998	92	45	,	,	PUNCT
fcis-6998	92	46	which	which	PRON
fcis-6998	92	47	can	can	AUX
fcis-6998	92	48	speed	speed	VERB
fcis-6998	92	49	up	up	ADP
fcis-6998	92	50	the	the	DET
fcis-6998	92	51	training	training	NOUN
fcis-6998	92	52	process	process	NOUN
fcis-6998	92	53	and	and	CCONJ
fcis-6998	92	54	improve	improve	VERB
fcis-6998	92	55	the	the	DET
fcis-6998	92	56	communication	communication	NOUN
fcis-6998	92	57	efficiency	efficiency	NOUN
fcis-6998	92	58	.	.	PUNCT
fcis-6998	93	1	as	as	SCONJ
fcis-6998	93	2	shown	show	VERB
fcis-6998	93	3	in	in	ADP
fcis-6998	93	4	fig	fig	NOUN
fcis-6998	93	5	2	2	NUM
fcis-6998	93	6	and	and	CCONJ
fcis-6998	93	7	fig	fig	NOUN
fcis-6998	93	8	3	3	NUM
fcis-6998	93	9	,	,	PUNCT
fcis-6998	93	10	the	the	DET
fcis-6998	93	11	overall	overall	ADJ
fcis-6998	93	12	observation	observation	NOUN
fcis-6998	93	13	of	of	ADP
fcis-6998	93	14	this	this	DET
fcis-6998	93	15	figure	figure	NOUN
fcis-6998	93	16	shows	show	VERB
fcis-6998	93	17	that	that	SCONJ
fcis-6998	93	18	with	with	ADP
fcis-6998	93	19	the	the	DET
fcis-6998	93	20	continuous	continuous	ADJ
fcis-6998	93	21	increase	increase	NOUN
fcis-6998	93	22	of	of	ADP
fcis-6998	93	23	training	training	NOUN
fcis-6998	93	24	rounds	round	NOUN
fcis-6998	93	25	,	,	PUNCT
fcis-6998	93	26	the	the	DET
fcis-6998	93	27	accuracy	accuracy	NOUN
fcis-6998	93	28	(	(	PUNCT
fcis-6998	93	29	acc	acc	PROPN
fcis-6998	93	30	)	)	PUNCT
fcis-6998	93	31	under	under	ADP
fcis-6998	93	32	different	different	ADJ
fcis-6998	93	33	parameter	parameter	NOUN
fcis-6998	93	34	conditions	condition	NOUN
fcis-6998	93	35	shows	show	VERB
fcis-6998	93	36	an	an	DET
fcis-6998	93	37	upward	upward	ADJ
fcis-6998	93	38	trend	trend	NOUN
fcis-6998	93	39	.	.	PUNCT
fcis-6998	94	1	in	in	ADP
fcis-6998	94	2	the	the	DET
fcis-6998	94	3	edge	edge	NOUN
fcis-6998	94	4	iid	iid	NOUN
fcis-6998	94	5	scene	scene	NOUN
fcis-6998	94	6	,	,	PUNCT
fcis-6998	94	7	no	no	ADV
fcis-6998	94	8	matter	matter	ADV
fcis-6998	94	9	how	how	SCONJ
fcis-6998	94	10	the	the	DET
fcis-6998	94	11	lines	line	NOUN
fcis-6998	94	12	of	of	ADP
fcis-6998	94	13	four	four	NUM
fcis-6998	94	14	colors	color	NOUN
fcis-6998	94	15	rise	rise	VERB
fcis-6998	94	16	,	,	PUNCT
fcis-6998	94	17	the	the	PRON
fcis-6998	94	18	smaller	small	ADJ
fcis-6998	94	19	the	the	DET
fcis-6998	94	20	model	model	NOUN
fcis-6998	94	21	accuracy	accuracy	NOUN
fcis-6998	94	22	is	be	AUX
fcis-6998	94	23	,	,	PUNCT
fcis-6998	94	24	the	the	PRON
fcis-6998	94	25	more	more	ADV
fcis-6998	94	26	accurate	accurate	ADJ
fcis-6998	94	27	it	it	PRON
fcis-6998	94	28	is	be	AUX
fcis-6998	94	29	.	.	PUNCT
fcis-6998	95	1	no	no	ADV
fcis-6998	95	2	matter	matter	ADV
fcis-6998	95	3	how	how	SCONJ
fcis-6998	95	4	it	it	PRON
fcis-6998	95	5	changes	change	VERB
fcis-6998	95	6	,	,	PUNCT
fcis-6998	95	7	as	as	ADV
fcis-6998	95	8	long	long	ADV
fcis-6998	95	9	as	as	SCONJ
fcis-6998	95	10	it	it	PRON
fcis-6998	95	11	is	be	AUX
fcis-6998	95	12	the	the	DET
fcis-6998	95	13	same	same	ADJ
fcis-6998	95	14	,	,	PUNCT
fcis-6998	95	15	the	the	DET
fcis-6998	95	16	accuracy	accuracy	NOUN
fcis-6998	95	17	curve	curve	NOUN
fcis-6998	95	18	of	of	ADP
fcis-6998	95	19	the	the	DET
fcis-6998	95	20	tested	test	VERB
fcis-6998	95	21	model	model	NOUN
fcis-6998	95	22	is	be	AUX
fcis-6998	95	23	basically	basically	ADV
fcis-6998	95	24	consistent	consistent	ADJ
fcis-6998	95	25	,	,	PUNCT
fcis-6998	95	26	which	which	PRON
fcis-6998	95	27	means	mean	VERB
fcis-6998	95	28	that	that	SCONJ
fcis-6998	95	29	increasing	increase	VERB
fcis-6998	95	30	the	the	DET
fcis-6998	95	31	communication	communication	NOUN
fcis-6998	95	32	frequency	frequency	NOUN
fcis-6998	95	33	between	between	ADP
fcis-6998	95	34	mec	mec	PROPN
fcis-6998	95	35	and	and	CCONJ
fcis-6998	95	36	ecs	ecs	PROPN
fcis-6998	95	37	will	will	AUX
fcis-6998	95	38	not	not	PART
fcis-6998	95	39	increase	increase	VERB
fcis-6998	95	40	the	the	DET
fcis-6998	95	41	training	training	NOUN
fcis-6998	95	42	duration	duration	NOUN
fcis-6998	95	43	.	.	PUNCT
fcis-6998	96	1	as	as	SCONJ
fcis-6998	96	2	shown	show	VERB
fcis-6998	96	3	in	in	ADP
fcis-6998	96	4	figure	figure	NOUN
fcis-6998	96	5	3	3	NUM
fcis-6998	96	6	,	,	PUNCT
fcis-6998	96	7	when=60	when=60	NOUN
fcis-6998	96	8	in	in	ADP
fcis-6998	96	9	the	the	DET
fcis-6998	96	10	edge	edge	NOUN
fcis-6998	96	11	niid	niid	PROPN
fcis-6998	96	12	scenario	scenario	PROPN
fcis-6998	96	13	,	,	PUNCT
fcis-6998	96	14	increasing	increase	VERB
fcis-6998	96	15	will	will	AUX
fcis-6998	96	16	delay	delay	VERB
fcis-6998	96	17	the	the	DET
fcis-6998	96	18	training	training	NOUN
fcis-6998	96	19	process	process	NOUN
fcis-6998	96	20	and	and	CCONJ
fcis-6998	96	21	affect	affect	VERB
fcis-6998	96	22	the	the	DET
fcis-6998	96	23	training	training	NOUN
fcis-6998	96	24	accuracy	accuracy	NOUN
fcis-6998	96	25	of	of	ADP
fcis-6998	96	26	the	the	DET
fcis-6998	96	27	model	model	NOUN
fcis-6998	96	28	.	.	PUNCT
fcis-6998	97	1	the	the	DET
fcis-6998	97	2	experimental	experimental	ADJ
fcis-6998	97	3	results	result	NOUN
fcis-6998	97	4	show	show	VERB
fcis-6998	97	5	that	that	SCONJ
fcis-6998	97	6	the	the	DET
fcis-6998	97	7	communication	communication	NOUN
fcis-6998	97	8	frequency	frequency	NOUN
fcis-6998	97	9	between	between	ADP
fcis-6998	97	10	the	the	DET
fcis-6998	97	11	vehicle	vehicle	NOUN
fcis-6998	97	12	and	and	CCONJ
fcis-6998	97	13	the	the	DET
fcis-6998	97	14	edge	edge	NOUN
fcis-6998	97	15	server	server	NOUN
fcis-6998	97	16	can	can	AUX
fcis-6998	97	17	be	be	AUX
fcis-6998	97	18	further	far	ADV
fcis-6998	97	19	improved	improve	VERB
fcis-6998	97	20	in	in	ADP
fcis-6998	97	21	order	order	NOUN
fcis-6998	97	22	to	to	PART
fcis-6998	97	23	reduce	reduce	VERB
fcis-6998	97	24	the	the	DET
fcis-6998	97	25	high	high	ADJ
fcis-6998	97	26	cost	cost	NOUN
fcis-6998	97	27	of	of	ADP
fcis-6998	97	28	traditional	traditional	ADJ
fcis-6998	97	29	direct	direct	ADJ
fcis-6998	97	30	communication	communication	NOUN
fcis-6998	97	31	with	with	ADP
fcis-6998	97	32	the	the	DET
fcis-6998	97	33	ecs	ecs	PROPN
fcis-6998	97	34	.	.	PUNCT
fcis-6998	98	1	the	the	DET
fcis-6998	98	2	training	training	NOUN
fcis-6998	98	3	duration	duration	NOUN
fcis-6998	98	4	of	of	ADP
fcis-6998	98	5	minist	minist	NOUN
fcis-6998	98	6	dataset	dataset	VERB
fcis-6998	98	7	under	under	ADP
fcis-6998	98	8	the	the	DET
fcis-6998	98	9	edge	edge	NOUN
fcis-6998	98	10	i	i	PROPN
fcis-6998	98	11	d	d	PROPN
fcis-6998	98	12	and	and	CCONJ
fcis-6998	98	13	edge	edge	VERB
fcis-6998	98	14	niid	niid	NOUN
fcis-6998	98	15	distribution	distribution	NOUN
fcis-6998	98	16	is	be	AUX
fcis-6998	98	17	shown	show	VERB
fcis-6998	98	18	in	in	ADP
fcis-6998	98	19	table	table	NOUN
fcis-6998	98	20	1	1	NUM
fcis-6998	98	21	.	.	PUNCT
fcis-6998	98	22	figure	figure	NOUN
fcis-6998	98	23	2	2	NUM
fcis-6998	98	24	.	.	PUNCT
fcis-6998	98	25	test	test	NOUN
fcis-6998	98	26	precision	precision	NOUN
fcis-6998	98	27	under	under	ADP
fcis-6998	98	28	minist	minist	PROPN
fcis-6998	98	29	data	data	PROPN
fcis-6998	98	30	set(iid	set(iid	PROPN
fcis-6998	98	31	)	)	PUNCT
fcis-6998	98	32	figure	figure	VERB
fcis-6998	98	33	3	3	NUM
fcis-6998	98	34	.	.	PUNCT
fcis-6998	98	35	test	test	NOUN
fcis-6998	98	36	precision	precision	NOUN
fcis-6998	98	37	under	under	ADP
fcis-6998	98	38	minist	minist	NOUN
fcis-6998	98	39	data	datum	NOUN
fcis-6998	98	40	set	set	VERB
fcis-6998	98	41	(	(	PUNCT
fcis-6998	98	42	niid	niid	NOUN
fcis-6998	98	43	)	)	PUNCT
fcis-6998	98	44	table	table	NOUN
fcis-6998	98	45	1	1	NUM
fcis-6998	98	46	.	.	PUNCT
fcis-6998	99	1	raining	rain	VERB
fcis-6998	99	2	duration	duration	NOUN
fcis-6998	99	3	of	of	ADP
fcis-6998	99	4	minist	minist	NOUN
fcis-6998	99	5	dataset	dataset	VERB
fcis-6998	99	6	under	under	ADP
fcis-6998	99	7	edge	edge	NOUN
fcis-6998	99	8	i	i	PROPN
fcis-6998	99	9	d	d	PROPN
fcis-6998	99	10	and	and	CCONJ
fcis-6998	99	11	edge	edge	VERB
fcis-6998	99	12	niid	niid	PROPN
fcis-6998	99	13	distribution	distribution	NOUN
fcis-6998	99	14	edge	edge	NOUN
fcis-6998	99	15	-	-	PUNCT
fcis-6998	99	16	iid	iid	NOUN
fcis-6998	99	17	edge	edge	NOUN
fcis-6998	99	18	-	-	PUNCT
fcis-6998	99	19	niid	niid	NOUN
fcis-6998	99	20	totalt	totalt	NOUN
fcis-6998	99	21	(	(	PUNCT
fcis-6998	99	22	s	s	NOUN
fcis-6998	99	23	)	)	PUNCT
fcis-6998	99	24	totalt	totalt	NOUN
fcis-6998	99	25	(	(	PUNCT
fcis-6998	99	26	s	s	NOUN
fcis-6998	99	27	)	)	PUNCT
fcis-6998	99	28	1	1	NUM
fcis-6998	100	1	=	=	SYM
fcis-6998	100	2	6	6	NUM
fcis-6998	100	3	,	,	PUNCT
fcis-6998	100	4	2	2	NUM
fcis-6998	100	5	=	=	SYM
fcis-6998	100	6	10	10	NUM
fcis-6998	100	7	4680	4680	NUM
fcis-6998	100	8	6120	6120	NUM
fcis-6998	100	9	1	1	NUM
fcis-6998	101	1	=	=	SYM
fcis-6998	101	2	15	15	NUM
fcis-6998	101	3	,	,	PUNCT
fcis-6998	101	4	2	2	NUM
fcis-6998	101	5	=	=	SYM
fcis-6998	101	6	4	4	NUM
fcis-6998	101	7	10140	10140	NUM
fcis-6998	101	8	11760	11760	NUM
fcis-6998	101	9	1	1	NUM
fcis-6998	101	10	=	=	SYM
fcis-6998	101	11	30	30	NUM
fcis-6998	101	12	,	,	PUNCT
fcis-6998	101	13	2	2	NUM
fcis-6998	101	14	=	=	SYM
fcis-6998	101	15	2	2	NUM
fcis-6998	101	16	13920	13920	NUM
fcis-6998	101	17	12360	12360	NUM
fcis-6998	101	18	1	1	NUM
fcis-6998	101	19	=	=	SYM
fcis-6998	101	20	60	60	NUM
fcis-6998	101	21	,	,	PUNCT
fcis-6998	101	22	2	2	NUM
fcis-6998	101	23	=	=	SYM
fcis-6998	101	24	1	1	NUM
fcis-6998	101	25	19680	19680	NUM
fcis-6998	101	26	28740	28740	NUM
fcis-6998	101	27	next	next	ADV
fcis-6998	101	28	,	,	PUNCT
fcis-6998	101	29	we	we	PRON
fcis-6998	101	30	study	study	VERB
fcis-6998	101	31	another	another	DET
fcis-6998	101	32	key	key	ADJ
fcis-6998	101	33	quantity	quantity	NOUN
fcis-6998	101	34	of	of	ADP
fcis-6998	101	35	the	the	DET
fcis-6998	101	36	three	three	NUM
fcis-6998	101	37	-	-	PUNCT
fcis-6998	101	38	layer	layer	NOUN
fcis-6998	101	39	federated	federated	ADJ
fcis-6998	101	40	learning	learning	NOUN
fcis-6998	101	41	system	system	NOUN
fcis-6998	101	42	,	,	PUNCT
fcis-6998	101	43	namely	namely	ADV
fcis-6998	101	44	the	the	DET
fcis-6998	101	45	training	training	NOUN
fcis-6998	101	46	duration	duration	NOUN
fcis-6998	101	47	of	of	ADP
fcis-6998	101	48	the	the	DET
fcis-6998	101	49	edge	edge	NOUN
fcis-6998	101	50	car	car	NOUN
fcis-6998	101	51	node	node	NOUN
fcis-6998	101	52	.	.	PUNCT
fcis-6998	102	1	in	in	ADP
fcis-6998	102	2	table	table	NOUN
fcis-6998	102	3	1	1	NUM
fcis-6998	102	4	,	,	PUNCT
fcis-6998	102	5	we	we	PRON
fcis-6998	102	6	compared	compare	VERB
fcis-6998	102	7	the	the	DET
fcis-6998	102	8	cloud	cloud	NOUN
fcis-6998	102	9	based	base	VERB
fcis-6998	102	10	federated	federated	ADJ
fcis-6998	102	11	learning	learning	NOUN
fcis-6998	102	12	of	of	ADP
fcis-6998	102	13	hierarchical	hierarchical	ADJ
fcis-6998	102	14	federated	federated	ADJ
fcis-6998	102	15	learning	learning	NOUN
fcis-6998	102	16	under	under	ADP
fcis-6998	102	17	different	different	ADJ
fcis-6998	102	18	scenarios	scenario	NOUN
fcis-6998	102	19	when	when	SCONJ
fcis-6998	102	20	the	the	DET
fcis-6998	102	21	data	datum	NOUN
fcis-6998	102	22	is	be	AUX
fcis-6998	102	23	fixed	fix	VERB
fcis-6998	102	24	.	.	PUNCT
fcis-6998	103	1	according	accord	VERB
fcis-6998	103	2	to	to	ADP
fcis-6998	103	3	the	the	DET
fcis-6998	103	4	data	datum	NOUN
fcis-6998	103	5	in	in	ADP
fcis-6998	103	6	table	table	NOUN
fcis-6998	103	7	1	1	NUM
fcis-6998	103	8	,	,	PUNCT
fcis-6998	103	9	we	we	PRON
fcis-6998	103	10	can	can	AUX
fcis-6998	103	11	clearly	clearly	ADV
fcis-6998	103	12	know	know	VERB
fcis-6998	103	13	that	that	SCONJ
fcis-6998	103	14	when	when	SCONJ
fcis-6998	103	15	the	the	DET
fcis-6998	103	16	training	training	NOUN
fcis-6998	103	17	duration	duration	NOUN
fcis-6998	103	18	reaches	reach	VERB
fcis-6998	103	19	a	a	DET
fcis-6998	103	20	certain	certain	ADJ
fcis-6998	103	21	accuracy	accuracy	NOUN
fcis-6998	103	22	,	,	PUNCT
fcis-6998	103	23	it	it	PRON
fcis-6998	103	24	decreases	decrease	VERB
fcis-6998	103	25	monotonously	monotonously	ADV
fcis-6998	103	26	.	.	PUNCT
fcis-6998	104	1	this	this	PRON
fcis-6998	104	2	shows	show	VERB
fcis-6998	104	3	that	that	SCONJ
fcis-6998	104	4	from	from	ADP
fcis-6998	104	5	the	the	DET
fcis-6998	104	6	perspective	perspective	NOUN
fcis-6998	104	7	of	of	ADP
fcis-6998	104	8	reducing	reduce	VERB
fcis-6998	104	9	communication	communication	NOUN
fcis-6998	104	10	costs	cost	NOUN
fcis-6998	104	11	,	,	PUNCT
fcis-6998	104	12	the	the	DET
fcis-6998	104	13	three	three	NUM
fcis-6998	104	14	-	-	PUNCT
fcis-6998	104	15	layer	layer	NOUN
fcis-6998	104	16	federated	federated	ADJ
fcis-6998	104	17	learning	learning	NOUN
fcis-6998	104	18	proposed	propose	VERB
fcis-6998	104	19	by	by	ADP
fcis-6998	104	20	us	we	PRON
fcis-6998	104	21	is	be	AUX
fcis-6998	104	22	better	well	ADJ
fcis-6998	104	23	than	than	ADP
fcis-6998	104	24	the	the	DET
fcis-6998	104	25	two	two	NUM
fcis-6998	104	26	-	-	PUNCT
fcis-6998	104	27	layer	layer	NOUN
fcis-6998	104	28	federated	federate	VERB
fcis-6998	104	29	learning	learning	NOUN
fcis-6998	104	30	based	base	VERB
fcis-6998	104	31	on	on	ADP
fcis-6998	104	32	cloud	cloud	NOUN
fcis-6998	104	33	.	.	PUNCT
fcis-6998	105	1	38	38	NUM
fcis-6998	105	2	5	5	NUM
fcis-6998	105	3	.	.	PUNCT
fcis-6998	106	1	summary	summary	NOUN
fcis-6998	106	2	this	this	DET
fcis-6998	106	3	paper	paper	NOUN
fcis-6998	106	4	proposes	propose	VERB
fcis-6998	106	5	a	a	DET
fcis-6998	106	6	three	three	NUM
fcis-6998	106	7	-	-	PUNCT
fcis-6998	106	8	layer	layer	NOUN
fcis-6998	106	9	federated	federate	VERB
fcis-6998	106	10	learning	learn	VERB
fcis-6998	106	11	system	system	NOUN
fcis-6998	106	12	of	of	ADP
fcis-6998	106	13	edge	edge	NOUN
fcis-6998	106	14	car	car	NOUN
fcis-6998	106	15	network	network	NOUN
fcis-6998	106	16	edge	edge	NOUN
fcis-6998	106	17	server	server	PROPN
fcis-6998	106	18	cloud	cloud	PROPN
fcis-6998	106	19	server	server	NOUN
fcis-6998	106	20	,	,	PUNCT
fcis-6998	106	21	which	which	PRON
fcis-6998	106	22	is	be	AUX
fcis-6998	106	23	supported	support	VERB
fcis-6998	106	24	by	by	ADP
fcis-6998	106	25	the	the	DET
fcis-6998	106	26	distributed	distribute	VERB
fcis-6998	106	27	algorithm	algorithm	NOUN
fcis-6998	106	28	hierfedprox	hierfedprox	ADJ
fcis-6998	106	29	.	.	PUNCT
fcis-6998	107	1	the	the	DET
fcis-6998	107	2	system	system	NOUN
fcis-6998	107	3	divides	divide	VERB
fcis-6998	107	4	the	the	DET
fcis-6998	107	5	local	local	ADJ
fcis-6998	107	6	update	update	NOUN
fcis-6998	107	7	into	into	ADP
fcis-6998	107	8	two	two	NUM
fcis-6998	107	9	parts	part	NOUN
fcis-6998	107	10	:	:	PUNCT
fcis-6998	107	11	the	the	DET
fcis-6998	107	12	edge	edge	NOUN
fcis-6998	107	13	car	car	NOUN
fcis-6998	107	14	node	node	NOUN
fcis-6998	107	15	and	and	CCONJ
fcis-6998	107	16	the	the	DET
fcis-6998	107	17	edge	edge	NOUN
fcis-6998	107	18	car	car	NOUN
fcis-6998	107	19	server	server	NOUN
fcis-6998	107	20	.	.	PUNCT
fcis-6998	108	1	the	the	DET
fcis-6998	108	2	model	model	NOUN
fcis-6998	108	3	aggregation	aggregation	NOUN
fcis-6998	108	4	is	be	AUX
fcis-6998	108	5	completed	complete	VERB
fcis-6998	108	6	on	on	ADP
fcis-6998	108	7	the	the	DET
fcis-6998	108	8	edge	edge	NOUN
fcis-6998	108	9	server	server	NOUN
fcis-6998	108	10	and	and	CCONJ
fcis-6998	108	11	the	the	DET
fcis-6998	108	12	cloud	cloud	NOUN
fcis-6998	108	13	server	server	NOUN
fcis-6998	108	14	.	.	PUNCT
fcis-6998	109	1	experiments	experiment	NOUN
fcis-6998	109	2	show	show	VERB
fcis-6998	109	3	that	that	SCONJ
fcis-6998	109	4	compared	compare	VERB
fcis-6998	109	5	with	with	ADP
fcis-6998	109	6	the	the	DET
fcis-6998	109	7	traditional	traditional	ADJ
fcis-6998	109	8	twolayer	twolayer	NOUN
fcis-6998	109	9	federated	federated	ADJ
fcis-6998	109	10	learning	learning	NOUN
fcis-6998	109	11	based	base	VERB
fcis-6998	109	12	on	on	ADP
fcis-6998	109	13	cloud	cloud	NOUN
fcis-6998	109	14	,	,	PUNCT
fcis-6998	109	15	it	it	PRON
fcis-6998	109	16	can	can	AUX
fcis-6998	109	17	reduce	reduce	VERB
fcis-6998	109	18	the	the	DET
fcis-6998	109	19	total	total	ADJ
fcis-6998	109	20	communication	communication	NOUN
fcis-6998	109	21	and	and	CCONJ
fcis-6998	109	22	computing	computing	NOUN
fcis-6998	109	23	costs	cost	NOUN
fcis-6998	109	24	at	at	ADP
fcis-6998	109	25	the	the	DET
fcis-6998	109	26	same	same	ADJ
fcis-6998	109	27	time	time	NOUN
fcis-6998	109	28	.	.	PUNCT
fcis-6998	110	1	in	in	ADP
fcis-6998	110	2	addition	addition	NOUN
fcis-6998	110	3	,	,	PUNCT
fcis-6998	110	4	it	it	PRON
fcis-6998	110	5	is	be	AUX
fcis-6998	110	6	also	also	ADV
fcis-6998	110	7	revealed	reveal	VERB
fcis-6998	110	8	that	that	SCONJ
fcis-6998	110	9	the	the	DET
fcis-6998	110	10	selection	selection	NOUN
fcis-6998	110	11	of	of	ADP
fcis-6998	110	12	different	different	ADJ
fcis-6998	110	13	key	key	ADJ
fcis-6998	110	14	parameters	parameter	NOUN
fcis-6998	110	15	in	in	ADP
fcis-6998	110	16	the	the	DET
fcis-6998	110	17	hierfedprox	hierfedprox	ADJ
fcis-6998	110	18	algorithm	algorithm	NOUN
fcis-6998	110	19	can	can	AUX
fcis-6998	110	20	achieve	achieve	VERB
fcis-6998	110	21	different	different	ADJ
fcis-6998	110	22	effects	effect	NOUN
fcis-6998	110	23	.	.	PUNCT
fcis-6998	111	1	references	reference	NOUN
fcis-6998	111	2	[	[	X
fcis-6998	111	3	1	1	NUM
fcis-6998	111	4	]	]	X
fcis-6998	111	5	blanco	blanco	PROPN
fcis-6998	111	6	b	b	PROPN
fcis-6998	111	7	,	,	PUNCT
fcis-6998	111	8	oscar	oscar	PROPN
fcis-6998	111	9	fajardo	fajardo	PROPN
fcis-6998	111	10	j	j	PROPN
fcis-6998	111	11	,	,	PUNCT
fcis-6998	111	12	giannoulakis	giannoulakis	PROPN
fcis-6998	112	1	i	i	PRON
fcis-6998	112	2	,	,	PUNCT
fcis-6998	112	3	et	et	PROPN
fcis-6998	112	4	al	al	PROPN
fcis-6998	112	5	.	.	PUNCT
fcis-6998	112	6	technology	technology	NOUN
fcis-6998	112	7	pillars	pillar	NOUN
fcis-6998	112	8	in	in	ADP
fcis-6998	112	9	the	the	DET
fcis-6998	112	10	architecture	architecture	NOUN
fcis-6998	112	11	of	of	ADP
fcis-6998	112	12	future	future	ADJ
fcis-6998	112	13	5	5	NUM
fcis-6998	112	14	g	g	NOUN
fcis-6998	112	15	mobile	mobile	ADJ
fcis-6998	112	16	networks	network	NOUN
fcis-6998	112	17	:	:	PUNCT
fcis-6998	112	18	nfv	nfv	NOUN
fcis-6998	112	19	,	,	PUNCT
fcis-6998	112	20	mec	mec	PROPN
fcis-6998	112	21	and	and	CCONJ
fcis-6998	112	22	sdn[j	sdn[j	PROPN
fcis-6998	112	23	]	]	PUNCT
fcis-6998	112	24	.	.	PUNCT
fcis-6998	113	1	computer	computer	NOUN
fcis-6998	113	2	standards	standard	NOUN
fcis-6998	113	3	and	and	CCONJ
fcis-6998	113	4	interfaces	interface	NOUN
fcis-6998	113	5	.	.	PUNCT
fcis-6998	114	1	2017(4):54	2017(4):54	NUM
fcis-6998	114	2	.	.	PUNCT
fcis-6998	115	1	[	[	X
fcis-6998	115	2	2	2	NUM
fcis-6998	115	3	]	]	X
fcis-6998	115	4	li	li	PROPN
fcis-6998	115	5	zuozhao	zuozhao	PROPN
fcis-6998	115	6	,	,	PUNCT
fcis-6998	115	7	liu	liu	PROPN
fcis-6998	115	8	jinxu	jinxu	PROPN
fcis-6998	115	9	,	,	PUNCT
fcis-6998	115	10	li	li	PROPN
fcis-6998	115	11	,	,	PUNCT
fcis-6998	115	12	et	et	PROPN
fcis-6998	115	13	al	al	PROPN
fcis-6998	115	14	application	application	NOUN
fcis-6998	115	15	of	of	ADP
fcis-6998	115	16	mobile	mobile	ADJ
fcis-6998	115	17	edge	edge	NOUN
fcis-6998	115	18	computing	compute	VERB
fcis-6998	115	19	in	in	ADP
fcis-6998	115	20	internet	internet	NOUN
fcis-6998	115	21	of	of	ADP
fcis-6998	115	22	vehicles	vehicle	NOUN
fcis-6998	115	23	[	[	X
fcis-6998	115	24	j	j	X
fcis-6998	115	25	]	]	X
fcis-6998	115	26	modern	modern	ADJ
fcis-6998	115	27	telecommunications	telecommunication	NOUN
fcis-6998	115	28	technology	technology	NOUN
fcis-6998	115	29	,	,	PUNCT
fcis-6998	115	30	2017	2017	NUM
fcis-6998	115	31	,	,	PUNCT
fcis-6998	115	32	47	47	NUM
fcis-6998	115	33	(	(	PUNCT
fcis-6998	115	34	3	3	NUM
fcis-6998	115	35	):	):	PUNCT
fcis-6998	115	36	5	5	NUM
fcis-6998	115	37	.	.	PUNCT
fcis-6998	116	1	[	[	X
fcis-6998	116	2	3	3	X
fcis-6998	116	3	]	]	X
fcis-6998	116	4	kekki	kekki	PROPN
fcis-6998	116	5	s	s	PROPN
fcis-6998	116	6	,	,	PUNCT
fcis-6998	116	7	featherstone	featherstone	PROPN
fcis-6998	116	8	w	w	PROPN
fcis-6998	116	9	,	,	PUNCT
fcis-6998	116	10	fang	fang	PROPN
fcis-6998	116	11	y	y	PROPN
fcis-6998	116	12	,	,	PUNCT
fcis-6998	116	13	et	et	PROPN
fcis-6998	116	14	al	al	PROPN
fcis-6998	116	15	.	.	PROPN
fcis-6998	116	16	mec	mec	PROPN
fcis-6998	116	17	in	in	ADP
fcis-6998	116	18	5	5	NUM
fcis-6998	116	19	g	g	NOUN
fcis-6998	116	20	networks	network	NOUN
fcis-6998	116	21	.	.	PUNCT
fcis-6998	117	1	etsi	etsi	PROPN
fcis-6998	117	2	white	white	PROPN
fcis-6998	117	3	paper	paper	PROPN
fcis-6998	117	4	,	,	PUNCT
fcis-6998	117	5	2018	2018	NUM
fcis-6998	117	6	,	,	PUNCT
fcis-6998	117	7	28(2018	28(2018	NUM
fcis-6998	117	8	):	):	PUNCT
fcis-6998	117	9	1	1	NUM
fcis-6998	117	10	-	-	SYM
fcis-6998	117	11	28	28	NUM
fcis-6998	117	12	.	.	PUNCT
fcis-6998	118	1	[	[	X
fcis-6998	118	2	4	4	X
fcis-6998	118	3	]	]	X
fcis-6998	118	4	fan	fan	PROPN
fcis-6998	118	5	y	y	PROPN
fcis-6998	118	6	,	,	PUNCT
fcis-6998	118	7	li	li	PROPN
fcis-6998	118	8	y	y	PROPN
fcis-6998	118	9	,	,	PUNCT
fcis-6998	118	10	zhan	zhan	PROPN
fcis-6998	118	11	m	m	PROPN
fcis-6998	118	12	,	,	PUNCT
fcis-6998	119	1	et	et	PROPN
fcis-6998	119	2	al	al	PROPN
fcis-6998	119	3	.	.	PUNCT
fcis-6998	119	4	iotdefender	iotdefender	PROPN
fcis-6998	119	5	:	:	PUNCT
fcis-6998	119	6	a	a	DET
fcis-6998	119	7	federated	federated	ADJ
fcis-6998	119	8	transfer	transfer	NOUN
fcis-6998	119	9	learning	learning	NOUN
fcis-6998	119	10	intrusion	intrusion	NOUN
fcis-6998	119	11	detection	detection	NOUN
fcis-6998	119	12	framework	framework	NOUN
fcis-6998	119	13	for	for	ADP
fcis-6998	119	14	5	5	NUM
fcis-6998	119	15	g	g	NOUN
fcis-6998	119	16	iot//	iot//	PROPN
fcis-6998	119	17	2020	2020	NUM
fcis-6998	119	18	ieee	ieee	PROPN
fcis-6998	119	19	14th	14th	ADJ
fcis-6998	119	20	international	international	ADJ
fcis-6998	119	21	conference	conference	NOUN
fcis-6998	119	22	on	on	ADP
fcis-6998	119	23	big	big	ADJ
fcis-6998	119	24	data	datum	NOUN
fcis-6998	119	25	science	science	NOUN
fcis-6998	119	26	and	and	CCONJ
fcis-6998	119	27	engineering	engineering	NOUN
fcis-6998	119	28	(	(	PUNCT
fcis-6998	119	29	bigdatase	bigdatase	NOUN
fcis-6998	119	30	)	)	PUNCT
fcis-6998	119	31	.	.	PUNCT
fcis-6998	120	1	ieee	ieee	PROPN
fcis-6998	120	2	,	,	PUNCT
fcis-6998	120	3	2020	2020	NUM
fcis-6998	120	4	.	.	PUNCT
fcis-6998	121	1	[	[	X
fcis-6998	121	2	5	5	NUM
fcis-6998	121	3	]	]	PUNCT
fcis-6998	121	4	ye	ye	PROPN
fcis-6998	121	5	y	y	PROPN
fcis-6998	121	6	,	,	PUNCT
fcis-6998	121	7	li	li	PROPN
fcis-6998	121	8	s	s	PROPN
fcis-6998	121	9	,	,	PUNCT
fcis-6998	121	10	liu	liu	PROPN
fcis-6998	121	11	f	f	PROPN
fcis-6998	121	12	,	,	PUNCT
fcis-6998	121	13	et	et	PROPN
fcis-6998	121	14	al	al	PROPN
fcis-6998	121	15	.	.	PROPN
fcis-6998	121	16	edgefed	edgefed	PROPN
fcis-6998	121	17	:	:	PUNCT
fcis-6998	121	18	optimized	optimize	VERB
fcis-6998	121	19	federated	federated	ADJ
fcis-6998	121	20	learning	learning	NOUN
fcis-6998	121	21	based	base	VERB
fcis-6998	121	22	on	on	ADP
fcis-6998	121	23	edge	edge	NOUN
fcis-6998	121	24	computing	computing	NOUN
fcis-6998	121	25	.	.	PUNCT
fcis-6998	122	1	ieee	ieee	NOUN
fcis-6998	122	2	access	access	NOUN
fcis-6998	122	3	,	,	PUNCT
fcis-6998	122	4	2020	2020	NUM
fcis-6998	122	5	,	,	PUNCT
fcis-6998	122	6	8:209191	8:209191	NUM
fcis-6998	122	7	-	-	SYM
fcis-6998	122	8	209198	209198	NUM
fcis-6998	122	9	.	.	PUNCT
fcis-6998	123	1	[	[	X
fcis-6998	123	2	6	6	NUM
fcis-6998	123	3	]	]	X
fcis-6998	123	4	zhou	zhou	PROPN
fcis-6998	123	5	s	s	PROPN
fcis-6998	123	6	,	,	PUNCT
fcis-6998	123	7	netalkar	netalkar	PROPN
fcis-6998	123	8	p	p	PROPN
fcis-6998	123	9	p	p	PROPN
fcis-6998	123	10	,	,	PUNCT
fcis-6998	123	11	chang	chang	PROPN
fcis-6998	123	12	y	y	PROPN
fcis-6998	123	13	,	,	PUNCT
fcis-6998	123	14	et	et	PROPN
fcis-6998	123	15	al	al	PROPN
fcis-6998	123	16	.	.	PUNCT
fcis-6998	124	1	the	the	DET
fcis-6998	124	2	mec	mec	PROPN
fcis-6998	124	3	-	-	PUNCT
fcis-6998	124	4	based	base	VERB
fcis-6998	124	5	architecture	architecture	NOUN
fcis-6998	124	6	design	design	NOUN
fcis-6998	124	7	for	for	ADP
fcis-6998	124	8	low	low	ADJ
fcis-6998	124	9	-	-	PUNCT
fcis-6998	124	10	latency	latency	NOUN
fcis-6998	124	11	and	and	CCONJ
fcis-6998	124	12	fast	fast	ADJ
fcis-6998	124	13	hand	hand	NOUN
fcis-6998	124	14	-	-	PUNCT
fcis-6998	124	15	off	off	NOUN
fcis-6998	124	16	vehicular	vehicular	ADJ
fcis-6998	124	17	networking//	networking//	ADJ
fcis-6998	124	18	ieee	ieee	NOUN
fcis-6998	124	19	vehicular	vehicular	ADJ
fcis-6998	124	20	technology	technology	NOUN
fcis-6998	124	21	conference	conference	NOUN
fcis-6998	124	22	(	(	PUNCT
fcis-6998	124	23	vtc	vtc	PROPN
fcis-6998	124	24	)	)	PUNCT
fcis-6998	124	25	.	.	PUNCT
fcis-6998	125	1	ieee	ieee	PROPN
fcis-6998	125	2	,	,	PUNCT
fcis-6998	125	3	2018	2018	NUM
fcis-6998	125	4	.	.	PUNCT
fcis-6998	126	1	[	[	X
fcis-6998	126	2	7	7	X
fcis-6998	126	3	]	]	X
fcis-6998	126	4	xiao	xiao	PROPN
fcis-6998	126	5	z	z	PROPN
fcis-6998	126	6	,	,	PUNCT
fcis-6998	126	7	dai	dai	PROPN
fcis-6998	126	8	x	x	PROPN
fcis-6998	126	9	,	,	PUNCT
fcis-6998	126	10	jiang	jiang	PROPN
fcis-6998	126	11	h	h	PROPN
fcis-6998	126	12	,	,	PUNCT
fcis-6998	126	13	et	et	PROPN
fcis-6998	126	14	al	al	PROPN
fcis-6998	126	15	.	.	PROPN
fcis-6998	126	16	vehicular	vehicular	ADJ
fcis-6998	126	17	task	task	NOUN
fcis-6998	126	18	offloading	offload	VERB
fcis-6998	126	19	via	via	ADP
fcis-6998	126	20	heat	heat	NOUN
fcis-6998	126	21	-	-	PUNCT
fcis-6998	126	22	aware	aware	ADJ
fcis-6998	126	23	mec	mec	PROPN
fcis-6998	126	24	cooperation	cooperation	NOUN
fcis-6998	126	25	using	use	VERB
fcis-6998	126	26	game	game	NOUN
fcis-6998	126	27	-	-	PUNCT
fcis-6998	126	28	theoretic	theoretic	NOUN
fcis-6998	126	29	method	method	NOUN
fcis-6998	126	30	.	.	PUNCT
fcis-6998	127	1	ieee	ieee	NOUN
fcis-6998	127	2	internet	internet	NOUN
fcis-6998	127	3	of	of	ADP
fcis-6998	127	4	things	thing	NOUN
fcis-6998	127	5	journal	journal	NOUN
fcis-6998	127	6	,	,	PUNCT
fcis-6998	127	7	2019	2019	NUM
fcis-6998	127	8	,	,	PUNCT
fcis-6998	127	9	7(3	7(3	NUM
fcis-6998	127	10	):	):	PUNCT
fcis-6998	127	11	2038	2038	NUM
fcis-6998	127	12	-	-	SYM
fcis-6998	127	13	2052	2052	NUM
fcis-6998	127	14	.	.	PUNCT
fcis-6998	128	1	[	[	X
fcis-6998	128	2	8	8	NUM
fcis-6998	128	3	]	]	X
fcis-6998	128	4	zhang	zhang	PROPN
fcis-6998	128	5	y	y	PROPN
fcis-6998	128	6	,	,	PUNCT
fcis-6998	128	7	lopez	lopez	PROPN
fcis-6998	128	8	j	j	PROPN
fcis-6998	128	9	,	,	PUNCT
fcis-6998	128	10	wang	wang	PROPN
fcis-6998	128	11	z.	z.	PROPN
fcis-6998	128	12	mobile	mobile	PROPN
fcis-6998	128	13	edge	edge	PROPN
fcis-6998	128	14	computing	compute	VERB
fcis-6998	128	15	for	for	ADP
fcis-6998	128	16	vehicular	vehicular	ADJ
fcis-6998	128	17	networks	network	NOUN
fcis-6998	128	18	[	[	X
fcis-6998	128	19	from	from	ADP
fcis-6998	128	20	the	the	DET
fcis-6998	128	21	guest	guest	NOUN
fcis-6998	128	22	editors	editor	NOUN
fcis-6998	128	23	]	]	PUNCT
fcis-6998	128	24	.	.	PUNCT
fcis-6998	129	1	ieee	ieee	PROPN
fcis-6998	129	2	vehicular	vehicular	ADJ
fcis-6998	129	3	technology	technology	NOUN
fcis-6998	129	4	magazine	magazine	NOUN
fcis-6998	129	5	,	,	PUNCT
fcis-6998	129	6	2019	2019	NUM
fcis-6998	129	7	,	,	PUNCT
fcis-6998	129	8	14(1	14(1	NUM
fcis-6998	129	9	):	):	PUNCT
fcis-6998	129	10	27	27	NUM
fcis-6998	129	11	-	-	SYM
fcis-6998	129	12	108	108	NUM
fcis-6998	129	13	.	.	PUNCT
fcis-6998	130	1	[	[	X
fcis-6998	130	2	9	9	NUM
fcis-6998	130	3	]	]	X
fcis-6998	130	4	wang	wang	PROPN
fcis-6998	130	5	h	h	PROPN
fcis-6998	130	6	,	,	PUNCT
fcis-6998	130	7	li	li	PROPN
fcis-6998	130	8	x	x	PROPN
fcis-6998	130	9	,	,	PUNCT
fcis-6998	130	10	ji	ji	PROPN
fcis-6998	130	11	h	h	PROPN
fcis-6998	130	12	,	,	PUNCT
fcis-6998	130	13	et	et	PROPN
fcis-6998	130	14	al	al	PROPN
fcis-6998	130	15	.	.	PUNCT
fcis-6998	130	16	dynamic	dynamic	ADJ
fcis-6998	130	17	offloading	offload	VERB
fcis-6998	130	18	scheduling	scheduling	NOUN
fcis-6998	130	19	scheme	scheme	NOUN
fcis-6998	130	20	for	for	ADP
fcis-6998	130	21	mec	mec	NOUN
fcis-6998	130	22	-	-	PUNCT
fcis-6998	130	23	enabled	enable	VERB
fcis-6998	130	24	vehicular	vehicular	ADJ
fcis-6998	130	25	networks//2018	networks//2018	PROPN
fcis-6998	130	26	ieee	ieee	PROPN
fcis-6998	130	27	/	/	SYM
fcis-6998	130	28	cic	cic	PROPN
fcis-6998	130	29	international	international	ADJ
fcis-6998	130	30	conference	conference	NOUN
fcis-6998	130	31	on	on	ADP
fcis-6998	130	32	communications	communication	NOUN
fcis-6998	130	33	in	in	ADP
fcis-6998	130	34	china	china	PROPN
fcis-6998	130	35	(	(	PUNCT
fcis-6998	130	36	iccc	iccc	PROPN
fcis-6998	130	37	workshops	workshop	NOUN
fcis-6998	130	38	)	)	PUNCT
fcis-6998	130	39	.	.	PUNCT
fcis-6998	131	1	ieee	ieee	NOUN
fcis-6998	131	2	,	,	PUNCT
fcis-6998	131	3	2018	2018	NUM
fcis-6998	131	4	:	:	PUNCT
fcis-6998	131	5	206	206	NUM
fcis-6998	131	6	-	-	SYM
fcis-6998	131	7	210	210	NUM
fcis-6998	131	8	.	.	PUNCT
fcis-6998	132	1	[	[	X
fcis-6998	132	2	10	10	NUM
fcis-6998	132	3	]	]	X
fcis-6998	132	4	zhou	zhou	PROPN
fcis-6998	132	5	chuanxin	chuanxin	PROPN
fcis-6998	132	6	,	,	PUNCT
fcis-6998	132	7	sun	sun	PROPN
fcis-6998	132	8	yi	yi	PROPN
fcis-6998	132	9	,	,	PUNCT
fcis-6998	132	10	wang	wang	PROPN
fcis-6998	132	11	degang	degang	PROPN
fcis-6998	132	12	,	,	PUNCT
fcis-6998	132	13	et	et	PROPN
fcis-6998	132	14	al	al	PROPN
fcis-6998	132	15	a	a	DET
fcis-6998	132	16	survey	survey	NOUN
fcis-6998	132	17	of	of	ADP
fcis-6998	132	18	federated	federated	ADJ
fcis-6998	132	19	learning	learning	NOUN
fcis-6998	132	20	research	research	NOUN
fcis-6998	132	21	journal	journal	NOUN
fcis-6998	132	22	of	of	ADP
fcis-6998	132	23	network	network	NOUN
fcis-6998	132	24	and	and	CCONJ
fcis-6998	132	25	information	information	NOUN
fcis-6998	132	26	security	security	NOUN
fcis-6998	132	27	.	.	PUNCT
fcis-6998	133	1	[	[	X
fcis-6998	133	2	11	11	NUM
fcis-6998	133	3	]	]	X
fcis-6998	133	4	yang	yang	PROPN
fcis-6998	133	5	q	q	PROPN
fcis-6998	133	6	,	,	PUNCT
fcis-6998	133	7	liu	liu	PROPN
fcis-6998	133	8	y	y	PROPN
fcis-6998	133	9	,	,	PUNCT
fcis-6998	133	10	chen	chen	PROPN
fcis-6998	133	11	t	t	PROPN
fcis-6998	133	12	,	,	PUNCT
fcis-6998	133	13	et	et	PROPN
fcis-6998	133	14	al	al	PROPN
fcis-6998	133	15	.	.	PROPN
fcis-6998	133	16	federated	federated	ADJ
fcis-6998	133	17	machine	machine	NOUN
fcis-6998	133	18	learning	learning	NOUN
fcis-6998	133	19	:	:	PUNCT
fcis-6998	133	20	concept	concept	NOUN
fcis-6998	133	21	and	and	CCONJ
fcis-6998	133	22	applications	application	NOUN
fcis-6998	133	23	.	.	PUNCT
fcis-6998	134	1	acm	acm	PROPN
fcis-6998	134	2	transactions	transaction	NOUN
fcis-6998	134	3	on	on	ADP
fcis-6998	134	4	intelligent	intelligent	ADJ
fcis-6998	134	5	systems	system	NOUN
fcis-6998	134	6	and	and	CCONJ
fcis-6998	134	7	technology	technology	NOUN
fcis-6998	134	8	,	,	PUNCT
fcis-6998	134	9	2019	2019	NUM
fcis-6998	134	10	,	,	PUNCT
fcis-6998	134	11	10(2):1	10(2):1	NUM
fcis-6998	134	12	-	-	SYM
fcis-6998	134	13	19	19	NUM
fcis-6998	134	14	.	.	PUNCT
fcis-6998	135	1	[	[	X
fcis-6998	135	2	12	12	NUM
fcis-6998	135	3	]	]	X
fcis-6998	135	4	yang	yang	PROPN
fcis-6998	135	5	q	q	PROPN
fcis-6998	135	6	,	,	PUNCT
fcis-6998	135	7	liu	liu	PROPN
fcis-6998	135	8	y	y	PROPN
fcis-6998	135	9	,	,	PUNCT
fcis-6998	135	10	chen	chen	PROPN
fcis-6998	135	11	t	t	PROPN
fcis-6998	135	12	,	,	PUNCT
fcis-6998	135	13	et	et	PROPN
fcis-6998	135	14	al	al	PROPN
fcis-6998	135	15	.	.	PROPN
fcis-6998	135	16	federated	federated	ADJ
fcis-6998	135	17	machine	machine	NOUN
fcis-6998	135	18	learning	learning	NOUN
fcis-6998	135	19	:	:	PUNCT
fcis-6998	135	20	concept	concept	NOUN
fcis-6998	135	21	and	and	CCONJ
fcis-6998	135	22	applications	application	NOUN
fcis-6998	135	23	.	.	PUNCT
fcis-6998	136	1	acm	acm	PROPN
fcis-6998	136	2	transactions	transaction	NOUN
fcis-6998	136	3	on	on	ADP
fcis-6998	136	4	intelligent	intelligent	ADJ
fcis-6998	136	5	systems	system	NOUN
fcis-6998	136	6	and	and	CCONJ
fcis-6998	136	7	technology	technology	NOUN
fcis-6998	136	8	,	,	PUNCT
fcis-6998	136	9	2019	2019	NUM
fcis-6998	136	10	,	,	PUNCT
fcis-6998	136	11	10(2):1	10(2):1	NUM
fcis-6998	136	12	-	-	SYM
fcis-6998	136	13	19	19	NUM
fcis-6998	136	14	.	.	PUNCT
fcis-6998	137	1	[	[	X
fcis-6998	137	2	13	13	NUM
fcis-6998	137	3	]	]	X
fcis-6998	137	4	lu	lu	PROPN
fcis-6998	137	5	y	y	PROPN
fcis-6998	137	6	,	,	PUNCT
fcis-6998	137	7	huang	huang	PROPN
fcis-6998	137	8	x	x	PROPN
fcis-6998	137	9	,	,	PUNCT
fcis-6998	137	10	dai	dai	PROPN
fcis-6998	137	11	y	y	PROPN
fcis-6998	137	12	,	,	PUNCT
fcis-6998	137	13	et	et	PROPN
fcis-6998	137	14	al	al	PROPN
fcis-6998	137	15	.	.	PUNCT
fcis-6998	138	1	differentially	differentially	ADV
fcis-6998	138	2	private	private	ADJ
fcis-6998	138	3	asynchronous	asynchronous	ADJ
fcis-6998	138	4	federated	federated	ADJ
fcis-6998	138	5	learning	learning	NOUN
fcis-6998	138	6	for	for	ADP
fcis-6998	138	7	mobile	mobile	ADJ
fcis-6998	138	8	edge	edge	NOUN
fcis-6998	138	9	computing	computing	NOUN
fcis-6998	138	10	in	in	ADP
fcis-6998	138	11	urban	urban	ADJ
fcis-6998	138	12	informatics	informatic	NOUN
fcis-6998	138	13	.	.	PUNCT
fcis-6998	139	1	ieee	ieee	NOUN
fcis-6998	139	2	transactions	transaction	NOUN
fcis-6998	139	3	on	on	ADP
fcis-6998	139	4	industrial	industrial	ADJ
fcis-6998	139	5	informatics	informatic	NOUN
fcis-6998	139	6	,	,	PUNCT
fcis-6998	139	7	2019	2019	NUM
fcis-6998	139	8	,	,	PUNCT
fcis-6998	139	9	16(3	16(3	NUM
fcis-6998	139	10	):	):	PUNCT
fcis-6998	139	11	2134	2134	NUM
fcis-6998	139	12	-	-	SYM
fcis-6998	139	13	2143	2143	NUM
fcis-6998	139	14	.	.	PUNCT
fcis-6998	140	1	[	[	X
fcis-6998	140	2	14	14	NUM
fcis-6998	140	3	]	]	X
fcis-6998	140	4	lu	lu	PROPN
fcis-6998	140	5	y	y	PROPN
fcis-6998	140	6	,	,	PUNCT
fcis-6998	140	7	huang	huang	PROPN
fcis-6998	140	8	x	x	PROPN
fcis-6998	140	9	,	,	PUNCT
fcis-6998	140	10	dai	dai	PROPN
fcis-6998	140	11	y	y	PROPN
fcis-6998	140	12	,	,	PUNCT
fcis-6998	140	13	et	et	PROPN
fcis-6998	140	14	al	al	PROPN
fcis-6998	140	15	.	.	PROPN
fcis-6998	140	16	federated	federated	ADJ
fcis-6998	140	17	learning	learning	NOUN
fcis-6998	140	18	for	for	ADP
fcis-6998	140	19	data	datum	NOUN
fcis-6998	140	20	privacy	privacy	NOUN
fcis-6998	140	21	preservation	preservation	NOUN
fcis-6998	140	22	in	in	ADP
fcis-6998	140	23	vehicular	vehicular	ADJ
fcis-6998	140	24	cyber	cyber	ADJ
fcis-6998	140	25	-	-	PUNCT
fcis-6998	140	26	physical	physical	ADJ
fcis-6998	140	27	systems	system	NOUN
fcis-6998	140	28	.	.	PUNCT
fcis-6998	141	1	ieee	ieee	NOUN
fcis-6998	141	2	network	network	NOUN
fcis-6998	141	3	,	,	PUNCT
fcis-6998	141	4	2020	2020	NUM
fcis-6998	141	5	,	,	PUNCT
fcis-6998	141	6	34(3	34(3	NUM
fcis-6998	141	7	):	):	PUNCT
fcis-6998	141	8	50	50	NUM
fcis-6998	141	9	-	-	SYM
fcis-6998	141	10	56	56	NUM
fcis-6998	141	11	.	.	PUNCT
fcis-6998	142	1	[	[	X
fcis-6998	142	2	15	15	NUM
fcis-6998	142	3	]	]	X
fcis-6998	142	4	liu	liu	PROPN
fcis-6998	142	5	l	l	PROPN
fcis-6998	142	6	,	,	PUNCT
fcis-6998	142	7	zhang	zhang	PROPN
fcis-6998	142	8	j	j	PROPN
fcis-6998	142	9	,	,	PUNCT
fcis-6998	142	10	song	song	NOUN
fcis-6998	142	11	s	s	PART
fcis-6998	142	12	h	h	NOUN
fcis-6998	142	13	,	,	PUNCT
fcis-6998	142	14	et	et	PROPN
fcis-6998	142	15	al	al	PROPN
fcis-6998	142	16	.	.	PROPN
fcis-6998	142	17	client	client	NOUN
fcis-6998	142	18	-	-	PUNCT
fcis-6998	142	19	edge	edge	NOUN
fcis-6998	142	20	-	-	PUNCT
fcis-6998	142	21	cloud	cloud	NOUN
fcis-6998	142	22	hierarchical	hierarchical	ADJ
fcis-6998	142	23	federated	federate	VERB
fcis-6998	142	24	learning//icc	learning//icc	PROPN
fcis-6998	142	25	2020	2020	NUM
fcis-6998	142	26	-	-	SYM
fcis-6998	142	27	2020	2020	NUM
fcis-6998	142	28	ieee	ieee	NOUN
fcis-6998	142	29	international	international	ADJ
fcis-6998	142	30	conference	conference	NOUN
fcis-6998	142	31	on	on	ADP
fcis-6998	142	32	communications	communication	NOUN
fcis-6998	142	33	(	(	PUNCT
fcis-6998	142	34	icc	icc	PROPN
fcis-6998	142	35	)	)	PUNCT
fcis-6998	142	36	.	.	PUNCT
fcis-6998	143	1	ieee	ieee	PROPN
fcis-6998	143	2	,	,	PUNCT
fcis-6998	143	3	2020	2020	NUM
fcis-6998	143	4	:	:	PUNCT
fcis-6998	143	5	1	1	NUM
fcis-6998	143	6	-	-	SYM
fcis-6998	143	7	6	6	NUM
fcis-6998	143	8	.	.	PUNCT
fcis-6998	144	1	[	[	X
fcis-6998	144	2	16	16	NUM
fcis-6998	144	3	]	]	X
fcis-6998	144	4	liu	liu	PROPN
fcis-6998	144	5	l	l	PROPN
fcis-6998	144	6	,	,	PUNCT
fcis-6998	144	7	zhang	zhang	PROPN
fcis-6998	144	8	j	j	PROPN
fcis-6998	144	9	,	,	PUNCT
fcis-6998	144	10	song	song	NOUN
fcis-6998	144	11	s	s	PART
fcis-6998	144	12	h	h	NOUN
fcis-6998	144	13	,	,	PUNCT
fcis-6998	144	14	et	et	PROPN
fcis-6998	144	15	al	al	PROPN
fcis-6998	144	16	.	.	PROPN
fcis-6998	144	17	edge	edge	PROPN
fcis-6998	144	18	-	-	PUNCT
fcis-6998	144	19	assisted	assist	VERB
fcis-6998	144	20	hierarchical	hierarchical	ADJ
fcis-6998	144	21	federated	federated	ADJ
fcis-6998	144	22	learning	learning	NOUN
fcis-6998	144	23	with	with	ADP
fcis-6998	144	24	non	non	ADJ
fcis-6998	144	25	-	-	ADJ
fcis-6998	144	26	iid	iid	ADJ
fcis-6998	144	27	data	datum	NOUN
fcis-6998	144	28	.	.	PUNCT
fcis-6998	145	1	arxiv	arxiv	PROPN
fcis-6998	145	2	preprint	preprint	VERB
fcis-6998	145	3	arxiv:1905.06641	arxiv:1905.06641	PROPN
fcis-6998	145	4	,	,	PUNCT
fcis-6998	145	5	2019	2019	NUM
fcis-6998	145	6	.	.	PUNCT