Computer Systems Science & Engineering DOI:10.32604/csse.2022.024749
Article

Analysis of Cognitive Radio for LTE and 5G Waveforms

Ramesh Ramamoorthy1, Himanshu Sharma2, A. Akilandeswari3, Nidhi Gour2, Arun Kumar4,* and Mehedi Masud5

1Centre for Artificial Intelligence, Chennai Institute of Technology, Chennai, India
2Department of Computer Science and Engineering, JECRC University, Jaipur, 303905, India
3Institute of ECE, Saveetha School of Engineering, Chennai, India
4Department of Electronics and Communication Engineering, JECRC University, Jaipur, 303905, India
5Department of Computer Science, College of Computers and Information Technology, Taif University, 11099, Saudi Arabia
*Corresponding Author: Arun Kumar. Email: arun.kumar1986@live.com
Received: 29 October 2021; Accepted: 13 December 2021

Abstract: Spectrum sensing is one of the major concerns in reaching an efficient Quality of service (QOS) in the advanced mobile communication system. The advanced engineering sciences such as 5G, device 2 device communications (D2D), Internet of things (IoT), MIMO require a large spectrum for better service. Orthogonal frequency division multiplexing (OFDM) is not a choice in advanced radio due to the Cyclic Prefix (CP), wastage of the spectrum, and so on. Hence, it is important to explore the spectral efficient advanced waveform techniques and combine a cognitive radio (CR) with the 5G waveform to sense the idle spectrum, which overcomes the spectrum issue. The demand for spectrum is ever increasing; however, spectrum is limited and is an acutely scarce resource. To alleviate the issue, techniques like Cognitive Radios (CR) have been devised. However, such techniques are non-standardized, and many variations of CR algorithms have been tried and tested. This paper details the several spectrum sensing methods tailored for CR. We explain the benefits, uniqueness, and drawbacks of the various techniques to provide a comprehensive review of the scene, including all recent and novel techniques of CR. Finally, we provided experimental results for the performance of the CR for key 5G and beyond modulation techniques to elaborate the dependency of the CR techniques for CR applications and provide a competitive review of their performance. Experiments show that the CR integrated with NOMA shows better performance as compared with existing techniques.

Keywords: CR; Energy detection; FBMC; NOMA; OFDM

1  Introduction

In wireless communication, spectrum is always a significant constraint, but still, a large bandwidth (more than two-thirds of the available) is wasted due to improper utilization, affecting the system’s QoS (Quality of Services). In the year 1990, Joseph Mitola presented a technique based on software-defined radio (SDR), which can sense the idle spectrum, known as Cognitive radio (CR) [1].

Spectrum sensing is a crucial method, which tries to utilize the spectrum efficiently. In the latest research, it is seen that 90% of the spectrum is not utilized, and there is a scarcity of spectrum due to the exponential growth of wireless applications and devices [2]. At the beginning of the 19th century, communication services of telephony, radio, and television have been regulated according to the model to provide service to the public on basic fair and condition [3]. The main motive of 5G communications is to wield large sums of data traffic and providing several applications like Industrial Internet of Things (IIOT), smart home, automation, D2D communication, and high data-rate. As a result, data consumption will increase by 30% and we need an advanced radio to process the various applications. The advanced radio needs high spectral efficiency to provide a respectable caliber of services. The major challenge in the channeling out of the spectrum sensing technique is to sense the idle spectrum in the presence and absence of primary users (PU). The quality of service (QoS) of advanced mobile communication depends on spectral efficiency, low detection delay, high data-rate, low peak power, and accessing large numbers of devices. The successful regularization of 5G depends on developing and designing an advanced modulation scheme. The regulation of radio spectrum has different characteristics, they are:

•   Licensed Spectrum

•   Unlicensed spectrum

•   Open spectrum

Over 160000 licensed users in India utilize amateur radio. Licenses are granted by the WPC (Wireless, Planning, and Coordination wing) of the Government of India [4]. Unlicensed spectrum is the freely available spectrum range for civilian usage. Although it is unregulated, it is severely congested due to many users sharing the same spectrum. Unlicensed bands operate on the 2.4 GHz ISM and 5 GHz UNII bands. The Office of Communication (Of-Com) in US, (2010) opened the television white space (TVWS) for secondary user (Unlicensed user) utilization. TVWS is a VHF/UHF band; it is a large portion of the RF spectrum that has become vacant after switching from analog to digital TV. The band can be operated in a cognitive approach without the primary analog TV user. The basic function of CR is to sense the idle spectrum of the primary user (PU) and allocate the idle spectrum to the secondary user (SU) without any disturbance and interference to PU [5]. There are two primary aims of the CR: first, it should not create and interference and degrade the performance of PU. Secondly, it should effectively locate the idle and unused spectrum bands to enhance the system’s throughput [6]. The role of spectrum sensing is vital for both PU and SU. While the SU benefits from gaining additional bandwidth, the PUs interference is kept within an acceptable limit. CR attains higher spectrum availability using dynamic spectrum access [7] by permitting unlicensed users (Secondary Users) to use the available bandwidth from licensed users (Primary Users) while preventing any interference to the Primary users’ transmissions [8]. CR decreases the inefficiencies caused by spectral congestion found in standard wireless environments by giving the opportunistic application of the frequency bands that are not congested by licensed users [9]. Therefore, white space in the spectrum can be used by unlicensed users without causing any significant interference [10]. Such efficient spectrum sharing methods allow users to coexist on the same frequencies without interfering with each other [11]. The uniqueness of CR lies in its capability to dynamically re-configure to optimal network usage. CR networks are specialized wireless networks, so they face more security attacks than traditional wireless and wired networks. The general security objectives of all wireless systems are privacy, integrity, availability, and access control [12]. The main goal of CR technology is to increase the throughput of the network and minimize the obstruction for primary users. CR can measure, sense, discover and be aware of radio channels individuality, accessibility of spectrum, and radio operational setting. There are three important sensing schemes in CR. Spectrum detection techniques such as Energy Detection (ED) [13], Matched Filter (MF) [14], and Cyclostationary Detection (CD) [15] have been proposed in the recent years. Energy detection is one of the most popular and straightforward techniques which can efficiently detect the spectrum at a high Signal to Noise Ratio (SNR). In ED, the energy of the received signal is estimated and compared with the predetermined threshold value. If the energy of the received signal is greater than the threshold value, then the PU is detected else not. However, its performance reduces at low SNR and noisy channels. Cyclostationary is utilized to sense the PU’s spectrum using the mean and autocorrelation function of the transmitted signal. The CR detects the idle spectrum based on the following hypothesis, where (H1) and (H0) indicate the presence and absence of PU [16].

{H0:Xj(t)=Nj(t).H1:Xj(t)=hjS(t)+Nj(t),j=1,…Nu} (1)

where Xj(t) is the jth SU, Nu is the number of Second Pu, S(t) is the PU signal, Nj(t) is noise, and hj is the channel’s gain between PU and SUs. Cyclostationary detection is used at significantly smaller SNR conditions. It is robust to noise and can carry out better than many sensing schemes like energy power, but it has significantly more computational complications [17]. The Cyclostationary spectrum detection method performs satisfactorily compared to other detection schemes because it has noise rejection capability. However, it also has some disadvantages, including spectral leakage of high amplitude signals, non-linearity, and high cost of operation [18]. The matched filter is a robust approach for sensing the idle spectrum of PU. The performance of the matched filter is efficient when the receiver knows the channel state information in advance. It is commonly utilized in radar, wherein we measure reflected signals, to detect the initially transmitted signal [19]. The schematic of the matched filter is given in Fig. 1.

Figure 1: Sensing

The filter input c(t) consists of a pulse signal a(t) corrupted by noise b(t), is shown by:

c(t)=a(t)+b(t),0≤t≤T (2)

where T is an arbitrary internal observation, the pulse signal a(t) may represent a binary symbol 1 or 0 in a digital communication system. The b(t) is a sample function of a white noise process of zero mean and power spectral density No2 . The function of the receiver is to detect the pulse signal a(t) in an optimum manner, given the received signal d(t). In this work; we implement the ED spectrum sensing method for the OFDM, and Non Orthogonal Multiple Access (NOMA) systems. To enhance the spectrum efficiency of the overall system, Cognitive Radio is integrated with different modulation techniques. Probability of detection (Pd) vs. Probability of false alarm (Pfa), Pd vs. Signal to Noise Ratio (SNR), and Bit Error Rate (BER) vs. SNR for each of the modulation frameworks are calculated and analyzed for ED-based CR. The main objectives of the projected work are as follows:

•   To implement a Cognitive radio for advanced waveforms and compare with OFDM structure.

•   In this work, the detection of the spectrum is possible in both the absence and presence of the primary user.

•   The different parameters such as Probability of detection (Pd), Probability of false alarm (Pfa), BER, and PAPR are estimated and compared for OFDM, FBMC, and NOMA waveforms.

Tab. 1 indicates the related literature published so for in Cognitive radio.

2  System Model

2.1 Energy Detection in OFDM System

The schematic representation of ED using the OFDM system is given in Fig. 2. It is implemented by using Inverse Fast Fourier Transform (IFFT), CP, and Fast Fourier Transform (FFT) at the transmitting and receiver terminal of the system. OFDM structure utilized a Cyclic prefix (CP) to overcome the Inter Symbol Interference (ISI), which results in a loss of spectrum. ED detection is applied to the OFDM structure to determine the status of the PU [30]. The energy of the received signal is estimated and compared with the threshold value, and the decision is made.

Figure 2: ED with OFDM

The OFDM symbols with N subcarriers can be written as:

Z=[Z0,Z1,...ZN−1]T (3)

The time-domain of OFDM symbols is obtained by IFFT:

x(n)=1N∑M=0N−1XMexp⁡(j6.28MLN) (4)

where n is the index of OFDM symbols and L is denotes the overlapping value. The energy of the received signal is estimated as [31]:

z(n)=1N∑n=1N|x(n)|2 (5)

Energy detection is the best scheme for detecting independently distributed signals of high SNR conditions, but it is unsuitable for detecting correlated signals. The hypothesis for Ed is given as [32]:

H0:z(n)=σ(n) (6)

H1:z(n)=x(n)+σ(n) (7)

where z(n) is the received signal, σ(n) is the noise variance, x(n) is the transmitted signal, H0 indicates the absence of PU, and H1 denotes the presence of PU.

2.2 Energy Detection in FBMC System

The schematic of FBMC is given in Fig. 3. FBMC is based on the multi-carrier technique and is considered one of the strong contenders for the 5G waveform. It is implemented by using an array of filters at the transmitter and receiver of the system [33].

Figure 3: ED with FBMC

Let us consider an FBMC signal with S sub-blocks is given as:

Zms=[Z0s,Z1s,...,ZM−1S]T (8)

where s=1,2,...,S and n denotes the sub-carriers. The FBMC signal can express in real and imaginary forms:

Zms=rms+jims (9)

The number of sub-carriers is applied to the group of filters:

Zms(t)=rmsf(t−sTs)+jimsf(t−Ts2−sTs) (10)

Ts is the duration of symbols, and f being the response of the filter. The time-domain estimation of FBMC signal is obtained by applying an IFFT, given as:

Zms(K)=1N∑m=0N−1zms(t)exp−j2πKmN (11)

The energy of the FBMC signal is estimated as:

z(n)=1N∑n=1N|Zms(K)|2 (12)

The hypothesis of Ed is estimated for the detection of the spectrum, given as:

H0:z(n)=σ(n) (13)

H1:z(n)=Zms(K)+σ(n)

where z(n) is the received signal, σ(n) is the noise variance, Zms(K) is the transmitted signal.

2.3 Energy Detection in NOMA System

The schematic of NOMA is given in Fig. 4. It is also considered one of the best contenders for 5G waveform. The distribution of resources based on the Super Coding algorithm (SC) and interference is mitigated by Successive Interference Cancellation (SIC) [34]. The distribution of resources is uniform for all users. Hence, it guarantees a maximum throughput of the system. However, complexity in the receiver framework is seen as one of the constraints of NOMA [35].

Figure 4: System model

Let us consider a NOMA signal with S subcarriers given as:

X=[X0,X1,...XS−1]T (14)

The response NOMA signal with Successive Interference cancellation (SIC), Super Coding (SC), filters, IFFT and FFT is give as:

z(n)=expj2πFct∑n=0S−1x(n)∗c(t−sT) (15)

where h(t−nT) denotes the characteristics of the filter. The energy of FBMC signal is estimated as [36]:

y(n)=1N∑n=1N|z(n)|2 (16)

The hypothesis of Ed is estimated for the detection of spectrum, given as:

H0:y(n)=σ(n) (17)

H1:y(n)=z(n)+σ(n) (18)

where y(n) is the received signal, σ(n) is the noise variance, z(n) is the transmitted signal.

3  Simulation Results

Matlab R21 is used in an Intel core i7 CPU platform as a simulation environment. The effect of cognitive operation for the key modulation techniques. To estimate the performance of ED-based CR, we consider the following specifications: Quadrature Amplitude Modulation (16-QAM), Rayleigh channel, 64-subcarriers, 600 symbols and, 64-length FFT. The probability of detection (Pd) performance of ED is shown in Fig. 5. It is seen that the Pd is maximum at the SNR of –3, 4, and 4 dB for NOMA, FBMC, and OFDM. Hence it shows that ED with NOMA gives better performance as compared with FBMC and OFDM. The PFA curves of waveforms with ED are shown in Fig. 6. It is observed that the ED with OFDM has a high probability of detecting noise as the desired signal. However, the false detection characteristics of NOMA and FBMC are similar and better than the OFDM. The PAPR performance without applying the reduction algorithms is shown in Fig. 7. At the Complementary Cumulative Distribution Function (CCDF) of 10-3, the PAPR of OFDM is 10 dB, FBMC is 8.8 dB, and NOMA is 7 dB. Hence, it is concluded that the NOMA achieves a gain of 1.8 and 3 dB as compared with FBMC and OFDM. The BER performance of ED-based CR with OFDM, FBMC, and NOMA is shown in Fig. 8. The BER of 10-5 is achieved at the SNR of 5.8 dB for NOMA, 7.6 dB for FBMC, and 9.6 dB for OFDM. Hence, it is concluded that the efficiency of ED-based CR for NOMA is better than that of OFDM and FBMC.

Figure 5: Pd performance of ED

Figure 6: ED with PFA

Figure 7: PAPR performance

Figure 8: BER performance

4  Conclusion

In this article, we propose ED-based Cognitive Radio for modern beyond 5G modulation like NOMA, OFDM, and FBMC waveforms. The main aim of the proposed work is to analyze the spectral efficiency and capacity of the different systems mentioned above. The simulation results reveal that the Probability of detection and false alarm performance of ED with NOMA is better and achieved a gain of 2.1 and 3.2 dB as compared with FBMC and OFDM. Further, the throughput of the CR for different waveforms is studied by estimating the BER and PAPR. It is seen that the ED with NOMA outperforms the existing waveforms. Thus, we expose NOMA to be a primary candidate for beyond 5G modulation, which increases the capacity of the system and results in low PARP compared to the current modulation scheme.

Acknowledgement: We would like to give special thanks to Taif University Research supporting Project Number (TURSP-2020/10), Taif University, Taif, Saudi Arabia.

Funding Statement: This research was supported by the Taif University Researchers supporting project (TURSP 2020/10), Taif university, Taif, Saudi Arabia.

Conflicts of Interest: The authors declare that they have no conflicts of interest to report regarding the present study.

References

 1.  J. Mitola and G. Q. Maguire, “Cognitive radio: Making software radios more personal,” IEEE Personal Communications, vol. 6, no. 4, pp. 13–18, 1999. [Google Scholar]

 2.  N. Muchandi and R. Khanai, “Cognitive radio spectrum sensing: A survey,” in 2016 Int. Conf. on Electrical, Electronics, and Optimization Techniques (ICEEOT), Lahore, Pakistan, pp. 3233–3323, 2016. [Google Scholar]

 3.  A. Kumar, S. Bharti and M. Gupta, “FBMC vs. OFDM: 5g mobile communication system,” International Journal of Systems, Control and Communications, vol. 10, no. 3, pp. 250–264, 2019. [Google Scholar]

 4.  J. E. Kasser, “Amateur radio: Past, present and future,” in Proc. of the 1995 Int. Conf. on 100 Years of Radio, London, UK, pp. 120–127, 1995. [Google Scholar]

 5.  S. Haykin, D. J. Thomson and J. H. Reed, “Spectrum sensing for cognitive radio,” Proc. of the IEEE, vol. 97, no. 5, pp. 849–877, 2009. [Google Scholar]

 6.  A. Kumar, M. K. Sharma, K. Sengar and S. Kumar, “NOMA based CR for QAM-64 and QAM-256,” Egyptian Informatics Journal, vol. 21, no. 2, pp. 67–71, 2020. [Google Scholar]

 7.  A. Kumar and P. N. Kumar, “OFDM system with cyclostationary feature detection spectrum sensing,” ICT Express, vol. 5, no. 1, pp. 21–25, 2019. [Google Scholar]

 8.  Y. Liang, K. Chen, G. Y. Li and P. Mahonen, “Cognitive radio networking and communications: An overview,” IEEE Transactions on Vehicular Technology, vol. 60, no. 7, pp. 3386–3407, 2011. [Google Scholar]

 9.  M. H. Islam, C. L. Koh, S. W. Oh, X. Qing, Y. Y. Lai et al., “Spectrum survey in Singapore: Occupancy measurements and analysis,” in 3rd Int. Conf. on Cognitive Radio Oriented Wireless Networks and Communications (CrownCom 2008), Singapore, pp. 1–7, 2008. [Google Scholar]

10. D. Datla, A. M. Wyglinski and G. J. Minden, “A spectrum surveying framework for dynamic spectrum access networks,” IEEE Transactions on Vehicular Technology, vol. 58, no. 8, pp. 4158–4168, 2009. [Google Scholar]

11. A. Goldsmith, S. A. Jafar, I. Maric and S. Srinivasa, “Breaking spectrum gridlock with cognitive radios: An information-theoretic perspective: An information theoretic perspective,” Proc. of the IEEE, vol. 97, no. 5, pp. 894–914, 2009. [Google Scholar]

12. K. Tumuluru, P. Wang and D. Niyato, “A Novel spectrum-scheduling scheme for multichannel cognitive radio network and performance analysis,” IEEE Transactions on Vehicular Technology, vol. 60, no. 4, pp. 1849–1858, 2011. [Google Scholar]

13. P. A. Navrátil, H. Childs, D. S. Fussell and C. Lin, “Exploring the spectrum of dynamic scheduling algorithms for scalable distributed-memoryray tracing,” IEEE Transactions on Visualization and Computer Graphics, vol. 20, no. 6, pp. 893–906, 2014. [Google Scholar]

14. P. Nandhakumar and A. Kumar, “Analysis of OFDM system with energy detection spectrum sensing,” Indian Journal of Science and Technology, vol. 9, no. 16, pp. 1–6, 2016. [Google Scholar]

15. S. A. Saberali and N. C. Beaulieu, “Matched-filter detection of the presence of MPSK signals,” in 2014 Int. Symp. on Information Theory and its Applications, Victoria, BC, Canada, pp. 85–89, 2014. [Google Scholar]

16. K. Kim, I. A. Akbar, K. K. Bae, J. Um, C. M. Spooner et al., “Cyclostationary approaches to signal detection and classification in cognitive radio,” in 2007 2nd IEEE Int. Symp. on New Frontiers in Dynamic Spectrum Access Networks, Dublin, Ireland, pp. 212–215, 2007. [Google Scholar]

17. R. S. Roberts, W. A. Brown and H. H. Loomis, “Computationally efficient algorithms for cyclic spectral analysis,” IEEE Signal Processing Magazine, vol. 8, no. 2, pp. 38–49, 1991. [Google Scholar]

18. D. Noguera, L. Biard and M. Laugeois, “Cyclostationarity detectors for cognitive radio: Architectural tradeoffs,” Journal of Wireless Communication Network, Vol. 2010, no. 1, pp. 1–8, 2010. [Google Scholar]

19. M. Ghozzi, M. Dohler, F. Marx and J. Palicot, “Cognitive radio: Methods for detection of free bands,” Comptes Rendus Physique, vol. 7, no. 7, pp. 794–804, 2006. [Google Scholar]

20. G. L. Turin, “An introduction to digital matched filters,” Proceedings of the IEEE, vol. 64, no. 7, pp. 1092–1112, 1976. [Google Scholar]

21. I. F. Akyildiz, W. Y. Lee, M. C. Vuran and S. Mohanty, “Next generation/dynamic spectrum access cognitive radio wireless network-A survey,” Computer Networks, vol. 5, no. 13, pp. 2127–2159, 2006. [Google Scholar]

22. C. Tu and B. Champagne, “Subspace-Based blind channel estimation for MIMO-OFDM systems with reduced time averaging,” IEEE Transactions on Vehicular Technology, vol. 59, no. 3, pp. 1539–1544, 2010. [Google Scholar]

23. J. Xuehua and C. Peijiang, “Research and Simulation of MIMO-OFDM Wireless Transmission System,” in 2009 Int. Forum on Information Technology and Applications, China, pp. 83–86, 2009. [Google Scholar]

24. Y. Zhou, Y. Wang, T. Wang and K. Zhang, “Iterative inter-cell interference coordination in MU-MIMO systems,” in Vehicular Technology and advancements Conf. or abbreviated as VTC Spring 2011 IEEE 73rd, Budapest, Hungary, pp. 1–5, 2011. [Google Scholar]

25. J. G. Andrews, S. Buzzi, W. Cho, V. S. Hanly, A. Lozano et al., “What will 5G be,” IEEE Journal on Selected Areas in Communication, vol. 32, no. 6, pp. 1065–1082, 2014. [Google Scholar]

26. M. O. Hasna, “Average BER of multihop transmission systems over fading channels,” in 10th IEEE Int. Conf. on Electronics, Circuits and Systems, 2003. ICECS 2003. Proc. of the 2003, Sharjah, United Arab Emirates, pp. 723–726, 2003. [Google Scholar]

27. O. J. Kwon and Y. H. Ha, “Multi-carrier PAPR reduction method using sub-optimal PTS with threshold,” IEEE Transactions on Broadcasting, vol. 49, no. 2, pp. 232– 236, 2003. [Google Scholar]

28. S. Srinu and S. Sab, “Co-operative wideband sensing based on cyclostationary features with multiple malicious user elimination,” International Journal of Electronics and Communication, vol. 68, no. 7, pp. 702–709, 2013. [Google Scholar]

29. Z. Quan, S. Cui, A. H. Sayed and H. V. Poor, “Optimal multiband joint detection for spectrum sensing in cognitive radio networks,” Transactions on Signal Processing, vol. 57, no. 3, pp. 1128–1140, 2009. [Google Scholar]

30. J. E. Salt and H. H. Nguyen, “Performance prediction for energy detection of unknown signals,” IEEE Transactions on Vehicular Technology, vol. 57, no. 6, pp. 3900–3904, 2008. [Google Scholar]

31. D. Na and K. Choi, “DFT spreading-based low PAPR FBMC with embedded side information,” IEEE Transactions on Communications, vol. 68, no. 3, pp. 1731–1745, 2020. [Google Scholar]

32. Z. He, L. Zhou, Y. Chen and X. Ling, “Low-complexity PTS scheme for PAPR reduction in FBMC-OQAM systems,” IEEE Communication Letter, vol. 22, no. 11, pp. 2322–2325, 2018. [Google Scholar]

33. A. Kumar and M. Gupta, “A review on activities of fifth generation mobile communication system,” Alexandria Engineering Journal, vol. 57, no. 2, pp. 1125–1135, 2018. [Google Scholar]

34. A. K. Bairagi, Md. S. Munir, M. Alsenwi, N. H. Tran, S. S. Alshamrani et al., “Coexistence mechanism between eMBB and uRLLC in 5g wireless networks,” IEEE Transactions on Communications, vol. 69, no. 3, pp. 1736–1749, 2021. [Google Scholar]

35. M. Mounir, Mb. M. El, S. Berra, G. S. Gaba and M. Masud, “A novel hybrid precoding-companding technique for peak-to-average power ratio reduction in 5G and beyond,” Sensors, vol. 21, no. 4, pp. 1410, 2021. [Google Scholar]

36. A. Kumar, S. Ambigapathy, M. Masud, E. S. Jaha, S. Chakravarty et al., “An efficient hybrid PAPR reduction for 5G NOMA-FBMC waveforms,” Computers, Materials & Continua, vol. 69, no. 3, pp. 2967–2981, 2021. [Google Scholar]

This work is licensed under a Creative Commons Attribution 4.0 International License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.