Compressive and non-compressive reliable wideband spectrum sensing at sub-Nyquist rates

Abstract

A significant challenge of cognitive radio (CR) is to monitor wide ranges of the radio spectrum and unveil vacant spectrum subbands, whilst operating at affordable data acquisition rates. In this paper, we evaluate a number of wideband spectrum sensing techniques that alleviate the data acquisition rate bottleneck by operating at remarkably low sub-Nyquist rates; thus dubbed sub-Nyquist spectrum sensing. These methods were independently developed and are rarely (if at all) compared. Among other deductions, the non-compressive sub-Nyquist approach is shown to have numerous benefits and delivers competitive detection performance compared with techniques based on the compressive sensing methodology.