An Energy Approximation Model Based on Restricted Isometry Property in Compressive Spectrum Sensing for Cognitive Radio

Abstract

Compressive Sensing (CS) has proved to be an effective technique in terms of spectrum sensing of Cognitive Radio (CR). However the signal reconstruction is costly and unnecessary. Through analysis of Restricted Isometry Property (RIP) of CS, it is feasible to detect the primary user directly through the compressed signal. In this paper, a novel Energy Approximation model based on Compressive Spectrum Sensing (EA-CSS) in cognitive radios is proposed. In contrast with the traditional compressive spectrum sensing scheme, this method does not need to reconstruct the original signal and has lower complexity algorithms at the same time. In the modeling, the observed signal energy and the original signal energy is nearly equal. The cooperative sensing based on this model is also improved and shows more robust to the noise compare with the common cooperative sensing scheme. Simulation results show that the proposed techniques have a reliable detection performance with small number of samples.