Low Complexity Sub-Nyquist Wideband Spectrum Sensing for Cognitive Radio

In order to provide more available spectrum and quickly switchover, wideband spectrum sensing (WSS) is desirable for cognitive radio. Sub-Nyquist sampling is crucial for WSS to reduce energy and computational costs, while the existing algorithms based on sub-Nyquist sampling either compute very complexly or perform poorly at low signal to noise ratio. In this paper, we propose a blind wideband spectrum sensing scheme based on a multi-coset sub-Nyquist sampling system. In contrast to the traditional subspace decomposition-based algorithms which first compute correlation matrix and then make eigenvalue decomposition, the proposed scheme directly uses the sample sequences to estimate the support set of active channels at sub-Nyquist rate, without up-sampling and reconstructing the original signal. In order to compute a suitable threshold value to decrease false alarm, we derive the effects of noises on compressed sensing systems through theoretical analysis. Numerical simulation results show that the proposed scheme outperforms traditional subspace decomposition-based algorithms in detection probability and computational complexity.