Wideband spectrum detection based on Compressed Sensing in cooperative cognitive radio networks

Compressed Sensing (CS) is a promising theory that has the power to reconstruct a certain signal from far fewer samples than conventional methods. Wideband detection is a challenge in Cognitive Radio (CR) networks because of its requirement for high sampling rate. Recent research shows that CS theory can be well applied to wideband detection with much lower sampling rates. In this paper, we propose a novel iterative algorithm for the noise-involved wideband detection in CR networks. In the proposed scheme, based on the different current detection results, the weights of the Weighted l₁ Minimization (WP1) are adjusted adaptively with the aim of improving the detection result in the next iteration. We also utilize M-out-of-N method in the fusion center to improve our detection result. Finally we introduce a metric which provides a better measurement for the detection performance. Simulation results prove our algorithm to be effective with lower sampling rate.