Entropy based compressive wideband spectrum sensing for small-scale primary users

For wideband cognitive radio networks, small-scale primary users (SSPUs) such as wireless microphones and mobile devices are generally difficult to be accurately sensed/detected due to the use of weak transmit power and noise uncertainty. Recently, compressive sensing (CS) technologies and edge detection methods (such as wavelet transform) are widely researched and combined to achieve the effective wideband spectrum sensing with sub-Nyquist sampling rate. When the SSPUs exist in wideband cognitive radio networks, however, there will occur two challenging points: low signal-to-noise ratio (S-NR) conditions and reduction of spectrum sparsity, which may lead to performance degradation of the traditional edge-based compressive spectrum detection methods. To overcome these difficulties, this paper proposes a novel robust compressive wideband spectrum sensing algorith-m by exploiting the information entropy of the power spectrum of SSPUs and noise signals. Theoretical analysis results show that the sensed signals´ power spectrum follows different statistical distributions when the SSPUs are present or not. Therefore, the proposed algorithm uses the information entropy as a test statistic to measure the varying of the statistical properties of the compressive reconstructed power spectrum of the sensed signals, and then make a decision on whether the primary users are present or not. Simulation results show the effectiveness of the proposed algorithm for detecting the SSPUs under low SNRs.