6___AITI#8504___143-154 Advances in Technology Innovation, vol. 7, no. 2, 2022, pp. 143-154 Low Access Latency and High Throughput for Full-Duplex Cognitive Radio Using Mobility Caching Placement Strategy Madhunala Srilatha 1,* , Anantha Bharathi 2 1 Department of Electronics and Communication Engineering, Vardhaman College of Engineering, Hyderabad, India 2 Department of Electronics and Communication Engineering, University College of Engineering, Osmania University, Hyderabad, India Received 18 September 2021; received in revised form 11 November 2021; accepted 12 November 2021 DOI: https://doi.org/10.46604/2022.8504 Abstract In a remote framework, the densely allotted radio range is not effectively utilized by an authorized primary user, which limits a secondary user to avoid disturbance to a primary user while utilizing a channel. To solve the shortage problem of accessible range, this study proposes a mobility caching placement strategy (MCPS) based sliding window scheme to optimize the latency and throughput of cognitive radio system. The performance parameters of the proposed system are analyzed using energy detection metrics by considering full-duplex communication. By the MATLAB analysis and the evaluation with and without self-interference suppression, the results demonstrate that the entrance inactivity of the MCPS based sliding window is diminished by a factor of 1.9 as contrasted with the existing full-duplex systems. With the improved latency and throughput, the proposed method can avoid the ruinous impacts of drawn-out self-residual suppression more effectively compared to the existing approaches. Keywords: full-duplex cognitive radio, latency, mobility caching placement strategy, self-residual suppression 1. Introduction One of the biggest difficulties in remote framework today is the shortage of accessible range, which can be resolved by using one of the best multiple access schemes, i.e., non-orthogonal multiple access (NOMA) [1]. Nonetheless, ongoing investigations have demonstrated that while the radio range is thickly allotted, it is frequently not intensely involved or utilized by an authorized primary user (PU) [2]. A recurrence coordinated cognitive radio (CR) system is proposed to exploit this circumstance while permitting an unlicensed secondary user (SU) to entrepreneurially reuse authorized recurrence groups without creating interference to the authorized user. One of the major necessities of such a framework is that SU ought not to produce destructive obstruction to PU [3-4]. Therefore, SU’s handset must be equipped for detecting the radio channel to decide whether PU is available or not by checking various performance parameters [5]. For range re-utilization, it requires low-inertness medium access for high-need users or high-need transmission. In this framework, a low latency-sensitive continuous transmission requires to be on hold for a split second as PU powerfully gets to the channel for getting its pressing message over. Such situations are of specific significance for the ongoing administration, for example, virtual/increased reality and independent vehicles. Therefore, it is required to get a latency of less than 1 msec to meet the forthcoming 5G systems [6]. Javed et al. [7] have given various methods to sense the PU spectrum with its pros and cons. Framework models that intermittently stop the SU transmission to detect the channel have been broadly proposed to recognize the beginning of the PU transmission and given various performance parameters’ comparison. These methodologies present visually impaired interims, where the SU framework is transmitting and becomes incapable of recognizing the beginning of the PU * Corresponding author. E-mail address: ma_srilatha2@yahoo.co.in Tel.: +91-9290885384 Advances in Technology Innovation, vol. 7, no. 2, 2022, pp. 143-154 transmission. Diminishing the interim between progressive detecting spaces will decrease the proficiency and throughput of the SU framework; however, this will enhance its range detecting abilities. To solve this problem, Khalid et al. [8] and Singh et al. [9] described the use of a full-duplex scheme that senses the spectrum and transmits the data simultaneously, and Alves et al. [10] provides protection to the user’s spectrum. Irio et al. [11] and Gaspard et al. [12] developed a method to expand the throughput of remote correspondence framework, where the remote terminal is permitted to transmit and sense at the same time in the same recurrence band. The throughput can also be increased by increasing the detection probability [13]. Tiwari et al. [14] proposed a cooperative medium access control (MAC) protocol called MLCP, which improves the throughput even more by enabling simultaneous transmission of PU and SU. Kumar et al. [15] proposed a clustering algorithm along with an energy efficient power allocation scheme due to which huge improvement in throughput is achieved. However, sustaining the improvement under imperfect self-interference cancellation is not addressed. Mokhtarzadeh et al. [16] proposed a method to increase throughput for a full-duplex CR system. Previous research has analyzed the exchange of detection which requires further improvement to detect proficiency. The best way to enhance the identification of PU is by utilizing full-duplex CR under self-residual suppression (SRS), which senses the channel for every �� test. This makes high access inertness as PU may begin the transmission whenever possible and can identify the nearness of PU more rapidly. Ayesha et al. [17] and Ayesha et al. [18] have described the alternative approaches of simultaneously sensing and transmission, using SRS for improving the trade-off between sensing the spectrum and SU throughput. However, the faultless SRS cannot be achieved below the noise floor. In the existing work, the performance analysis of spectrum sensing is done by using various schemes and protocols, e.g., multi-objective modified grey wolf optimization algorithm [19], NOMA network under power splitting based simultaneous wireless information and power transfer (SWIPT) [20], NOMA based mobile edge computing with imperfect channel state information (CSI) [21], decode-and-forward (DF) and amplify-and-forward (AF) techniques in NOMA based cognitive radio network (CRN) [22], signal segmentation [23], neural network based multilayer perception model [24], and the channel allocation scheme based on greedy algorithm [25]. Cache technology is used to even improve the performance of parameters of spectrum sensing which includes various strategies: cache capacity allocation, cache replacement, cache utilization, and cache placement [26]. In the work of Zheng et al. [27] and Xia et al. [28], the performance improvement of parameters is analyzed including energy consumption optimization, system transmission performance, and latency. In this study, a mobility caching placement strategy (MCPS) based sliding window scheme is proposed. This method significantly reduces access latency for the full-duplex system under SRS to protect PU. Additionally, it improves the SU throughput which maximizes the data transmission. This study emphasizes the issue of compromising the performance parameters of spectrum sensing (such as throughput and access latency) in the full-duplex CR under SRS. With different schemes, e.g., half-duplex, full-duplex using slotted window, and full-duplex using sliding window, the performance of parameters is evaluated with and without SRS. To optimize the sensing parameters, a MCPS based sliding window is proposed. Compared to the existing methods, the proposed method will exchange more information bits by taking decisions based on each sample, resulting in a quick detection of PU, in turn lowering the latency, and improving the SU throughput. The study is organized as follows. The system model and frame structure of the existing methods and the proposed method is described in section 2. In section 3, the MCPS based sliding window scheme is discussed. The performance analysis of latency and throughput is analyzed in section 4. Analytical and simulation results are presented in section 5. Finally, section 6 concludes the study. 2. System Model While SU is using the spectrum of PU, no interference in the PU network should occur. To ensure this, SU needs to accurately identify the white spaces for the transmission and should vacate and opt for the subsequent band when PU appears. The 144 Advances in Technology Innovation, vol. 7, no. 2, 2022, pp. 143-154 first scenario is considered hypothesis H1. When PU is using the spectrum, the signal received at SU is expressed in Eq. (1). The second scenario is considered hypothesis H0. While PU is not using the spectrum, the received signal can be expressed in Eq. (2). ( ) ( ) ( )S n X n u n= + (1) ( ) ( )S n u n= (2) In the above equation, ���� is the signal which needs to be identified, ���� is the Gaussian noise which is additive white in nature, and n is the sample index. The parameters of performance using detection algorithm will be analyzed based on two probabilities: the detection probability (Pd) with hypothesis H1 for identifying the presence of licensed user activity and the false alarm probability (Pf) with hypothesis H0 for detecting the licensed user when the licensed user is actually not present. PU is effectively protected when the probability of detection is high. In another scenario, SU can use the frequency band recurrently when the false alarm probability is low and is available. To guarantee better performance of the system, the detection probability shall be as much as possible, and the false alarm probability shall be as minimum as possible. CR system schemes are half-duplex and full-duplex, which are addressed in this section. The presence of detection of PU is done by using a metric M, which is compared with a given threshold � for both schemes. PU is present if �, otherwise PU is absent and SU can use the channel. Energy detection is the frequently used metric for sensing the spectrum [29]. ( ) 2 1 1 s N ns M S n N = = ∑ (3) where ���� is the SU received signal. If the PU information is known prior, other detection methods can be used. Using decision metric, the detection probability Pd and the false alarm probability Pf can be evaluated with the help of a hypothesis test [7]. ( )0/rfP P M H= > γ (4) ( )1/rdP P M H= < γ (5) The CR system frame structure is shown in Fig. 1. Initially, the CR system has to sense the spectrum and determine the status of PU. If the frequency band of PU is detected as idle, then the SU transmitter will make use of the band to transmit information to the SU receiver for the frame duration. If PU starts transmission, then SU has to cease its transmission, to avoid interference to PU. Again, in the next frame, SU will access the frequency band and the process repeats. Fig. 1 Frame structure of CR system under SRS 145 Advances in Technology Innovation, vol. 7, no. 2, 2022, pp. 143-154 2.1. Half-duplex CR system In this scheme, spectrum sensing has to be scheduled sequentially with transmission, as shown in Fig. 2. The frame contains two parts: � �� sample window for transmission and �� sample window for sensing the frequency band. �� samples are used to know the status of PU using Eq. (1) and thus decisions will be made for every � sample. It is clear that under the half-duplex scheme, SU has to cease the transmission each time the spectrum sensing is established, which leads to the higher access latency of PU and the reduction of SU throughput by a factor of ���� � . Fig. 2 Frame structure of half-duplex CR system 2.2. Simultaneous detection and transmission with slotted window scheme The frame structure of simultaneous detection and transmission with a slotted window using SRS is shown in Fig. 3 [16]. It contains a single frame, in which both the sensing and transmission will be performed simultaneously to lead to the improvement in detection probability and the decisions made for every ��. In this scheme, as there is no separate sensing slot, the SU throughput will not reduce, and SU can detect the status of PU immediately which leads to a reduction of latency. Fig. 3 Frame structure of slotted window full-duplex CR system 2.3. Sliding window scheme for simultaneous detection and transmission The frame structure of simultaneous detection and transmission with a sliding window using SRS is shown in Fig. 4, where the decisions will be made for every sample compared to the slotted window (for every ��). This scheme can be actualized effortlessly in computerized equipment employing a first-in first-out (FIFO) cushion. It is imperative to take a note that progressive choice measurements are not free as just a single new sample is included (and one is expelled). Fig. 4 Frame structure of sliding window full-duplex CR system 2.4. Quadrature phase shift keying The major advantages of digital modulation techniques are the improved capacity to transmit huge data with high immunity and the easy detection of its distinct transmission state at the receiver in a noisy medium. While converting analogue signals to digital signals, a trade-off is always made due to the loss of some information in the quantization process. Proper selection of the digital modulation technique is crucial especially when the bandwidth and time slots are limited in 146 Advances in Technology Innovation, vol. 7, no. 2, 2022, pp. 143-154 uplink-downlink transmission. The efficacy of a modulation technique is restrained in terms of its bandwidth and power efficiency. The capability of a modulation technique in accommodating data within a limited bandwidth is regarded as bandwidth efficiency whereas the ability to preserve the bit error probability of the digital message at low power levels is treated as the power efficiency of a modulation technique. With regards to the digital modulation schemes, for the same number of bits transmitted and received, binary phase shift keying (BPSK) modulation has less bit error rate and bandwidth for different values of signal-to-noise ratio (SNR), as compared to binary amplitude shift keying (BASK) and binary frequency shift keying (BFSK) modulation. To detect the message at the output, minimum SNR and low bit error rate with higher performance is required. With low power consumption, BPSK modulation technique provides the best performance even for less SNR. Bandwidth can be saved even more in M-ary phase shift keying (M-PSK) modulation schemes by maintaining symbol rate constantly and increasing the number of bits per symbol. For example, in quadrature phase shift keying (QPSK), one of the 4 possible phase shifts per symbol can be transmitted rather than transmitting only 2 possible phase shifts so that 2 bits per symbol (i.e., for the same bandwidth with twice the information) can be transmitted [30]. Also, with the same energy efficiency, one can achieve double spectral efficiency. Higher compound M-PSK modulation schemes can also be possible (8-PSK, 32- PSK, 64- PSK, etc.) to transmit more bits per symbol and in turn achieve higher bit rate. However, this employs higher vulnerability to noise as the symbols together get the closer and increased bit errors. In order to make a trade-off between bit rate and probability of error, QPSK is opted in the proposed system. 3. MCPS Based Sliding Window Scheme Caching techniques are attracted expressively since they can decrease backhaul traffic and also improve performance metrics. Caching placement strategies can be of uncoded or coded type. Both are used for optimizing the performance of parameters. The MCPS scheme allows decisions to be taken on a multiple sample transfer at a time and increases the concurrent transmission and sensing using intra interference suppression, which checks the presence of PU continuously in a spectrum band and gives a perfect position. The operation of MCPS is shown in Fig. 5 and the flowchart is shown in Fig. 6. The number of input samples and coding rate is calculated by considering SNR, number of symbols present at the transmitter, and threshold value using Eqs. (6) and (7). ( ) ( )1 ..... 2 x bs bs N R SNR N SNR N= × × + + (6) / ( 10 )tx SNR N i R H Thr N= × + × (7) where N represents the number of samples, H is the analysis of H equation, SNR represents the signal-to-noise ratio, Rx is the receiver, Thr represents the threshold value, Ri represents the coding rate, and Ntx represents the number of symbols at the transmitter. A signal is passed through the QPSK modulator, then the modulated data is analyzed, processed using Eqs. (8)-(10), and then transmitted through the channel. ( ) ( ) ( ) ( )1 i bs bs iipQ R w L N k w L N R k = × − × × × − ×φ φ (8) ( ) ( ) bs tx w L ipQ w L N r = × × (9) 147 Advances in Technology Innovation, vol. 7, no. 2, 2022, pp. 143-154 ( ) ( )ipQ w s w L ifft= × (10) where ���� represents the phase of the signal, ���� is the line width, ipQ represents the quadrature input, rtx represents the random information at transmitter, and ��s� is the sample width. Parameter optimization is done by using Eqs. (11)-(17). ( ) ( ) ( ) ( ) 2 s i tx txT bs bs w L QP NP c R N N j w L r w L N= × × × − × × × × (11) ( ) ( ) ( ) ( )iTT R w s QP ifft QP R w L = × × (12) ( ) ( ) tx T TSQP r M QP conj QP w s = × ×  ∑ (13) ( ) ( )( )v p txT R T RP Q Max QP conj QP r = × −   (14) ( )( ) ( ) ( ) tx TSRN T R r M QP conj QP w s = × −∑ (15) SQP SRN M M Gain = (16) ( )1010 logMSE Gain= × (17) Fig. 5 Operation of MCPS Fig. 6 Flowchart of MCPS 148 Advances in Technology Innovation, vol. 7, no. 2, 2022, pp. 143-154 4. Performance Analysis 4.1. Analysis of latency and throughput in MCPS based sliding window The impact of detecting time on the throughput is examined in this section. In each edge of the frame structure span TF, SU starts sensing the channel for a period of sensing time (��), and if the channel is free (which indicates PU is not using the channel), then SU utilizes the channel (spectrum) for information transmission. On the other hand, if PU wants to use the channel, then SU has to evacuate the channel and use another channel that is free for transmission without creating interference to PU. If the length of the corresponding edge of PU is TPRF, then the throughput of SU that could be transmitted when PU is not utilizing the spectrum is TPRF + TF. The throughput (Th) of SU is given by: ( )1sF cf F T t Th P P T −= − (18) At the point, when PU has an exponential on-off movement display, with the mean spans of “on” period signified by β, then Pc is given by: exp PRF Fc T T P       + = − β (19) Expecting that the casing of span TPRF, TF, and false caution likelihood has been settled, then the standardized throughput of SU is: ( )1 expsF PRF Ff F T t T T Th P T       − + = − − β (20) For the case of H0 (i.e., PU is available but wrongly detected), the probability density function (PDF) of the SU received signal (S) is P0(x), and the probability of false alarms Pf is: ( ) ( )0 0/fP P S H P x dx ∞ = > = ∫ ε ε (21) The detection threshold Ɛ is calculated as: ( )2 1 1u f sQ P f− ε = σ τ +  (22) On the other hand, when PU is using the channel (for the case of H1), the pdf of the SU received signal (S) is P1(x), and the detection probability Pd is: ( ) ( )1 1/dP S H P x dx ∞ = > = ∫ ε ε (23) The latency (L) of the system is calculated as: ( )         ∩∩ ++         ∩ +         += − c i cc i iccc DDD D PN DD D PN D D PNDPNL 12121 3 3 1 2 211 ... ..... (24) where Ni is the number of samples when PU begins with i th decision point, Di denotes the event when PU is detected at i th decision once transmission starts, and ��� represents the complementary event when PU is not detected during ith decision. 149 Advances in Technology Innovation, vol. 7, no. 2, 2022, pp. 143-154 4.2. Algorithm The algorithm of MCPS based on latency and throughput optimization is described in Table 1, where the input parameters need to be specified first and then the Eqs. (8)-(10) are analyzed. Furthermore, the optimization of latency and throughput are calculated based on the analytical values. Table 1 Algorithm of MCPS Input: ��, ���, ���, ��, ��, � !, �!# Output: Return the estimated data in the QPSK domain [var] Begin For i = 1 to $!�% &' &( j = 1 to $!�% &' &( )*ℎ,� , �. = ���0!, 1!, ��, ��� For k = 1…………. $!�% &' using 8-10 in var �� Parameter Optimization Learn (Latency Variance) Δ3� Update (var name) 43� using � → �� + 1� or �� 1� Throughput = ∑ �9:�;:���3;:��<=> �? = @(AB% C C D� ��EFGHI End 5. Analytical and Simulation Results This section explains the analytical and simulation results of the proposed MCPS technique under perfect SRS and imperfect SRS, as compared with the existing methods. Simulation results show the improvement in performance parameters by using the proposed technique. Simulation parameters of the MCPS scheme are shown in Table 2. Table 2 Simulation parameters Parameter Value Channel length 128 Number of bits 10 5 Number of sub-carriers 128 Number of transmitters 2 Number of receivers 2 Fast Fourier transform size 64 Sensing window size 16 5.1. Perfect self-interference suppression Fig. 7 depicts the substantial improvement of latency and throughput with the MCPS based sliding full-duplex method compared to the existing methods under perfect SRS. Since half-duplex allows either sensing or transmission at a time and cannot perform both tasks at a time, the latency of this method is almost half of a SU frame length and throughput is also very less. Unlike half-duplex, the slotted full-duplex method allows both sensing and transmission at a time, and the latency of this method has been improved for a duration of ��/2 compared to half-duplex since, for each �� duration, there is a provision of taking decisions. As the decisions are taken for every sample, the sliding full-duplex method can detect the usage of the channel by PU quickly. Furthermore, compared to the sliding full-duplex, the latency of MCPS is improved by a factor of 0.98; compared to the slotted full-duplex, the MCPS latency is improved by a factor of 3.7; compared to half-duplex, the MCPS latency is improved by a factor of 20. Fig. 8 illustrates the different quantiles of optimization parameter trade-off for the proposed MCPS technique and the existing sliding and slotted techniques. The performance of parameters depends on self-residual, noise, and signal realizations. The results show the improvement of the proposed technique compared to the existing techniques. 150 Advances in Technology Innovation, vol. 7, no. 2, 2022, pp. 143-154 Fig. 7 Latency versus throughput comparison for all the 4 methods under perfect SRS Fig. 8 Quantiles of optimization parameter trade-off under perfect SRS 5.2. Imperfect self-interference suppression Fig. 9 depicts latency improvement for MCPS with an increase in SRS, measured with respect to the noise floor which is calculated by considering throughput to 0.9. When compared with the existing methods, the slotted full-duplex method has higher latency for all values of self-interference, but the proposed method can withstand the improvement of self-interference by a 3.2 factor. Fig. 10 depicts the throughput improvement for MCPS with increasing SRS. The MCPS method has improved the throughput under the PU protection. Table 3 shows the comparison of the parameters of the sliding full-duplex and MCPS methods. Fig. 9 Latency versus SRS Fig. 10 Throughput with increasing SRS for the latency of 16 samples Table 3 Comparison of parameters Parameters Sliding window full-duplex MCPS No. of transmitters (Tx) 2 2 No. of receivers (Rx) 2 2 Channel length 64 Up to 128 Number of bits 10 4 10 5 Channel size 52 52 Max latency 0.90 0.79 Throughput 0.78 0.89 Error rate 0.44 0.42 SNR 0.97 0.99 151 Advances in Technology Innovation, vol. 7, no. 2, 2022, pp. 143-154 6. Conclusions and Future Scope This study investigated CR from the point of view of ensuring the essential and high priority users under perfect SRS and imperfect SRS by diminishing the latency of the framework. By inferring systematic recipes for latency and throughput of the existing methods, the proposed method is beneficial in terms of its capacity to make choices at each sample. Additionally, the results of simulation show that the proposed sensing framework can achieve a huge improvement in latency and throughput under SRS, which is diminished by a factor of 1.9 contrasted with the current full-duplex system. The performance parameters can be further extended by using vehicular ad-hoc networks, which provide cooperation in CR and share the channel status while switching between the channels when PU exists. Furthermore, the parameters can be extended by using a testbed that tests the performance and practicability of CR framework. List of Acronyms NOMA Non-orthogonal multiple access SNR Signal-to-noise ratio PU Primary user BASK Binary amplitude shift keying SU Secondary user BFSK Binary frequency shift keying CR Cognitive radio M-PSK M-ary phase shift keying MAC Medium access control Rx Receiver SRS Self-residual suppression Ntx Number of symbols at transmitter SWIPT Simultaneous wireless information and power transfer QPSK Quadrature phase shift keying CSI Channel state information ���� The phase of the signal DF Decode-and-forward ���� Line width AF Amplify-and-forward ipQ Quadrature input CRN Cognitive radio network rtx Random information at transmitter MCPS Mobility caching placement strategy ��s� Sample width Pd Probability of detection MSE Mean square error Pf Probability of false alarm TF Frame structure span FIFO First-in first-out Ɛ Detection threshold BPSK Binary phase shift keying Th Throughput of SU �� Sensing time Β Mean span of “on” period �� Transmission time TPRF Length of the corresponding edge of PU ���� Detected signal PDF Probability density function ���� Additive white Gaussian noise L Latency M Decision metric Ni Number of samples at i th decision point � Threshold Di Event when PU detected at ith decision ���� Received signal ��� Complimentary event when PU is not detected during ith decision N Number of input samples Conflicts of Interest The authors declare no conflict of interest. References [1] A. Kumar and K. Kumar, “Multiple Access Schemes for Cognitive Radio Networks: A Survey,” Physical Communication, vol. 38, 100953, February 2020. [2] Federal Communications Commission, “Spectrum Policy Task Force Report”, Federal Communications Commission, Technical Report FCC 02-155, November 15, 2002. 152 Advances in Technology Innovation, vol. 7, no. 2, 2022, pp. 143-154 [3] J. Tiwari, A. Prakash, and R. Tripathi, “A Novel Cooperative MAC Protocol for Safety Applications in Cognitive Radio Enabled Vehicular Ad-Hoc Networks,” Vehicular Communications, vol. 29, 100336, June 2021. [4] O. Salameh, H. Bruneel, and S. Wittevrongel, “Performance Evaluation of Cognitive Radio Networks with Imperfect Spectrum Sensing and Bursty Primary User Traffic,” Mathematical Problems in Engineering, vol. 2020, 4102046, 2020. [5] A. El Rharras, M. Saber, A. Chehri, R. Saadane, N. Hakem, and G. Jeon, “Optimization of Spectrum Utilization Parameters in Cognitive Radio Using Genetic Algorithm,” Procedia Computer Science, vol. 176, pp. 2466-2475, October 2020. [6] B. Kodavati and M. Ramarakula, “Latency Based Re-Enforcement Learning over Cognitive Software Defined 5G Networks,” 7th International Conference on Advanced Computing and Communication Systems, pp. 1848-1853, March 2021. [7] J. N. Javed, M. Khalil, and A. Shabbir, “A Survey on Cognitive Radio Spectrum Sensing: Classifications and Performance Comparison,” International Conference on Innovative Computing, pp. 1-8, November 2019. [8] W. Khalid and H. Yu, “Sensing and Utilization of Spectrum with Cooperation Interference for Full-Duplex Cognitive Radio Networks,” 11th International Conference on Ubiquitous and Future Networks, pp. 598-600, July 2019. [9] V. Singh, A. Gadre, and S. Kumar, “Full-Duplex Radios: Are We There Yet?” Proc. of 19th ACM Workshop on Hot Topics in Networks, pp. 117-124, November 2020. [10] H. Alves, T. Riihonen, and H. A. Suraweera, Full-Duplex Communications for Future Wireless Networks, Singapore: Springer, 2020. [11] L. Irio, A. T. Abusabah, and R. Oliveira, “Residual Self-Interference Estimation in In-Band Full-Duplex Wireless Systems,” International Wireless Communications and Mobile Computing, pp. 1130-1134, June 2020. [12] G. Gaspard and D. S. Kim, “Optimal Sensing and Interference Suppression in 5G Cognitive Radio Networks,” Journal of Communications, vol. 15, no. 4, pp. 303-308, April 2020. [13] C. M. Wu, Y. C. Kao, and K. F. Chang, “A Multichannel MAC Protocol for IoT-Enabled Cognitive Radio Ad Hoc Networks,” Advances in Technology Innovation, vol. 5, no. 1, pp. 45-55, December 2019. [14] J. Tiwari, A. Prakash, and R. Tripathi, “A Multichannel Link-Layer Cooperation Protocol (MLCP) for Cognitive Radio Ad Hoc Network,” International Conference on VLSI, Communication, and Signal Processing, pp. 191-200, October 2019. [15] A. Kumar and K. Kumar, “Energy-Efficient Resource Optimization Using Game Theory in Hybrid NOMA Assisted Cognitive Radio Networks,” Physical Communication, vol. 47, 101382, August 2021. [16] H. Mokhtarzadeh, A. Taherpour, A. Taherpour, and S. Gazor, “Throughput Maximization in Energy Limited Full-Duplex Cognitive Radio Networks,” IEEE Transactions on Communications, vol. 67, no. 8, pp. 5287-5296, August 2019. [17] A. Ayesha and S. M. Chaudhry, “Self-Interference Cancellation for Full-Duplex Radio Transceivers Using Extended Kalman Filter,” National Academy Science Letters, vol. 43, no. 7, pp. 631-634, February 2020. [18] A. Ayesha, M. Rahman, A. Haider, and S. M. Chaudhry, “On Self-Interference Cancellation and Non-Idealities Suppression in Full-Duplex Radio Transceivers,” Mathematics, vol. 9, no. 12, 1434, June 2021. [19] G. Eappen and T. Shankar, “Multi-Objective Modified Grey Wolf Optimization Algorithm for Efficient Spectrum Sensing in the Cognitive Radio Network,” Arabian Journal for Science and Engineering, vol. 46, no. 4, pp. 3115-3145, November 2020. [20] V. Aswathi and A. V. Babu, “Full/Half Duplex Cooperative Relaying NOMA Network under Power Splitting Based SWIPT: Performance Analysis and Optimization,” Physical Communications, vol. 46, 101335, June 2021. [21] H. Jiang, Y. Wang, X. Yue, and X. Li, “Performance Analysis of NOMA-Based Mobile Edge Computing with Imperfect CSI,” EURASIP Journal on Wireless Communications and Networking, vol. 2020, 138, July 2020. [22] A. Kumar and K. Kumar, “Relay Sharing with DF and AF Techniques in NOMA Assisted Cognitive Radio Networks,” Physical Communication, vol. 42, 101143, October 2020. [23] K. Kirubahini, J. J. Triphena, P. G. S. Velmurugan, and S. J. Thiruvengadam, “Optimal Spectrum Sensing in Cognitive Radio Systems Using Signal Segmentation Algorithm,” International Conference on Wireless Communications, Signal Processing, and Networking, pp. 118-121, August 2020. [24] S. DasMahapatra, S. Patnaik, S. N. Sharan, and M. Gupta, “Performance Analysis of Prediction-Based Spectrum Sensing for Cognitive Radio Networks,” 2nd International Conference on Intelligent Communication and Computational Techniques, pp. 271-274, September 2019. [25] R. Li and P. Zhu, “Spectrum Allocation Strategies Based on QoS in Cognitive Vehicle Networks,” IEEE Access, vol. 8, pp. 99922-99933, May 2020. 153 Advances in Technology Innovation, vol. 7, no. 2, 2022, pp. 143-154 [26] L. Li, G. Zhao, and R. S. Blum, “A Survey of Caching Techniques in Cellular Networks: Research Issues and Challenges in Content Placement and Delivery Strategies,” IEEE Communications Surveys and Tutorials, vol. 20, no. 3, pp. 1710-1732, March 2018. [27] X. Zheng, G. Wang, and Q. Zhao, “A Cache Placement Strategy with Energy Consumption Optimization in Information-Centric Networking,” Future Internet, vol. 11, no. 3, 64, March 2019. [28] J. Xia, F. Zhou, X. Lai, H. Zhang, H. Chen, Q. Yang, et al., “Cache Aided Decode and Forward Relaying Networks: From the Spatial View,” Wireless Communications and Mobile Computing, vol. 2018, 5963584, April 2018. [29] H. Shehata and T. Khattab, “Energy Detection Spectrum Sensing in Full-Duplex Cognitive Radio: The Practical Case of Rician RSI,” IEEE Transactions on Communications, vol. 67, no. 9, pp. 6544-6555, September 2019. [30] N. Fatima, S. A. Siddiqui, and A. Ahmad, “Comparative Performance Analysis of Digital Modulation Schemes with Digital Audio Transmission through AWGN Channel,” International Conference on Power Electronics, Control, and Automation, pp. 1-3, November 2019. Copyright© by the authors. Licensee TAETI, Taiwan. This article is an open-access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY-NC) license (https://creativecommons.org/licenses/by-nc/4.0/). 154