Performance Metrics, Sampling Schemes, and Detection Algorithms for Wideband Spectrum Sensing

In this paper, we study the problem of wideband spectrum sensing for cognitive radio networks by partitioning it into four fundamental elements: system modeling, performance metrics, sampling schemes, and detection algorithms. Each element can potentially couple the individual channels so that designs for wideband spectrum sensing should consider the four elements jointly. We propose the p-sparse model for the primary occupancies and study three uniform sampling schemes for wideband spectrum sensing, specifically, partial-band Nyquist sampling, sequential narrowband Nyquist sampling, and integer undersampling. We suggest new performance metrics more appropriate for wideband spectrum sensing, specifically, the probability of insufficient spectrum opportunity and the probability of excessive interference opportunity. We also develop detection algorithms that effectively tradeoff these metrics. Our results indicate that for performance metrics that couple the individual channels, multichannel detection algorithms have a significant advantage over channel-by-channel detection algorithms even for wideband Nyquist sampling. Furthermore, integer undersampling, which corresponds to the simplest sub-Nyquist sampling scheme, along with suitable detection algorithms exhibits appealing detection performance in the regime that better protects the primary system.