Adaptive Compressive Spectrum Sensing for Wideband Cognitive Radios

Abstract

This letter presents an adaptive spectrum sensing algorithm that detects\nwideband spectrum using sub-Nyquist sampling rates. By taking advantage of\ncompressed sensing (CS), the proposed algorithm reconstructs the wideband\nspectrum from compressed samples. Furthermore, an l2 norm validation approach\nis proposed that enables cognitive radios (CRs) to automatically terminate the\nsignal acquisition once the current spectral recovery is satisfactory, leading\nto enhanced CR throughput. Numerical results show that the proposed algorithm\ncan not only shorten the spectrum sensing interval, but also improve the\nthroughput of wideband CRs.\n