Exploiting Correlation in Distributed Cooperative Compressive Wideband Spectrum Sensing

In distributed cooperative compressive wideband (WB) spectrum sensing (SS), two secondary users (SUs) can sense a wideband spectrum individually and report their information to the fusion center (FC) where joint decoding and decision is made. Since the same WB spectrum is concerned, the two quantized measurement vectors received at the FC are naturally correlated. This paper presents a Slepian-Wolf coding-based scheme to exploit this correlation. Specifically, we show that for a given number of total reporting bits, a larger number of measurements can be collected that contribute to better fusion results in terms of both mean squared error (MSE) in spectrum reconstruction and receiver operating characteristic (ROC) in spectrum detection.