Performance analysis of sub-Nyquist sampling for Wideband spectrum sensing in cognitive radio

Abstract

Spectrum sensing is a key function of cognitive radio (CR) to identify the vacant frequency bands (which is also referred to as spectrum holes). The future technology of CR networks should be capable to scan wideband frequencies to increase spectrum utilization. To reduce high sampling rate for sampling wideband signal, Modulated Wideband Converter (MWC) is used for data acquisition. In this paper, MWC is also used to detect vacant spectrum of wideband signal modeled as multiband signal. Performance of the system is evaluated through simulation using Monte Carlo method. Performance metrics such as probability of detection (Pd), probability of false alarm (Pf) are examined in various SNR with sparsity level is 6/195 and 10/195. Numerical results show that spectrum sensing with sub-Nyquist sampling using MWC gives good detection performance. Other then SNR value, Sparsity level of the signal also have contribution signal detection performance. Perfect recovered support is achieved at SNR = 1 dB for the case with sparsity level = 6/195. When sparsity level = 10/195, Pd = 1 is never achieved.