Cooperative Sensing of Wideband Cognitive Radio: A Multiple-Hypothesis-Testing Approach

Abstract

A low-complexity decision fusion method is proposed for cooperative spectrum sensing (SS) of wideband cognitive radio (CR) using multiple-hypothesis testing. To maintain the quality of SS in multiple channels, performance indices of false-sensing rate and false-ignorance rate are defined for the Benjamini-Hochberg (BH) procedure to control the false-alarm ratio (FAR) and missing ratio (MR) of SS, respectively. Extended from these results, a double BH procedure is further studied in an attempt to simultaneously attain a low FAR and MR. In addition to identifying the availability of channels, a signal strength-estimation scheme that features multiple-hypothesis testing is also proposed to classify the level of the SNR for cooperative SS. Simulation results show that the double BH procedure can suppress both the FAR and the MR at a high SNR while maintaining a considerable FAR in the low-SNR regime. As such, the system can enjoy throughput enhancement by allowing more active cognitive access at a high SNR and still prevent introducing excessive multiple-access interference to the primary users when their SNRs are weak. On the other hand, for SNR classification, simulations also show that the SNR can be classified within one level, plus or minus the true strength with more than a 96% chance. This allows the system to employ more generic protocols such as concurrent transmissions for cognitive access.