Detection of Spectrum in Cognitive Radio Based on a Hybrid Approach

Abstract

Recent years have seen a sharp increase in the number of wireless devices, which has led to a shortage of radio spectrum due to static spectrum allocation. The licensed spectrum bands, however, are not being used to their full potential, according to numerous studies on static allocation. Spectrum shortage and underuse have been problems that have been thought to be overcome via cognitive radio. To find spectrum holes, spectrum sensing (SS) is a crucial component of cognitive radio. To detect the presence or absence of primary user signals, many conventional spectrum sensing techniques such as Energy Detection (ED), Matched filter Detection (MFD), and Cyclo-stationary Feature Detection (CFD) have been developed. There are numerous difficulties in performance detection for these techniques. For instance, high susceptibility to noise uncertainty, high sensing time and need for prior knowledge of the primary user's signal are limitations of ED, CFD, and MFD respectively. In this thesis, a hybrid SS method consists of ED, MF, and CFD are proposed to exploit their merits and overcome their challenges. The proposed hybrid method aims to improve the detection performance of the existing hybrid technique over the Additive White Gaussian Noise (AWGN) channel and under the Rayleigh fading channel. Also, a performance comparison among conventional SS techniques, conventional hybrid based on ED and CFD, hybrid based on ED and MFD, and MFD and CFD SS over the AWGN channel are performed. The performance of the proposed hybrid approach is additionally enhanced by the cooperative detection methods. Parameters including the probability of detection, probability of false alarm, probability of miss detection, receiver operating characteristic (ROC) curve, and complementary receiver operating characteristic (CROC) curve are used to assess the effectiveness of SS. According to simulation results, the proposed hybrid technique significantly outperforms about 1.5 dB, 3.5 dB, and 5.5dB in SNR wall when compared to the hybrid method from MFD and CFD, ED and MFD, and conventional hybrid based on ED and CFD, respectively under AWGN. Additionally, under Rayleigh fading channel, the proposed technique achieves 1 dB, 1.5 dB, and 5.6 dB respectively. In contrast, a cooperative proposed hybrid technique involving two users shows a 6 dB and 4.5 dB performance improvement over single node proposed hybrid techniques under AWGN and Rayleigh fading channel respectively.