Wideband spectrum detection using compressed sampling under fading channel environments

Efficient spectrum detection is essential to Cognitive Radio systems. In this paper, a novel spectrum detection method based on Bayesian compressed sampling is proposed. Firstly, the sparsity of the received signal under fading channel is studied. Then, Bayesian compressed sampling(BCS) is introduced to the spectrum detection for wideband cognitive radio system, and the spectrum components can be reconstructed adaptively through relevance vector machine. The simulation results show that the proposed method have good performance on spectrum reconstruction quality in fading channel environments and is robust to channel noise as well, meanwhile the signal can be acquired at sub-Nyquist sampling rate.