Multi-rate Sub-Nyquist Spectrum Sensing in Cognitive Radios

Wideband spectrum sensing is becoming increasingly important to cognitive radio (CR) systems for exploiting spectral opportunities. This paper introduces a novel multi-rate sub-Nyquist spectrum sensing (MS) system that implements cooperative wideband spectrum sensing in a CR network. MS can detect the wideband spectrum using partial measurements without reconstructing the full frequency spectrum. Sub-Nyquist sampling rates are adopted in sampling channels for wrapping the frequency spectrum onto itself. This significantly reduces sensing requirements of CR. The effects of sub-Nyquist sampling are considered, and the performance of multi-channel sub-Nyquist samplings is analyzed. To improve its detection performance, sub-Nyquist sampling rates are chosen to be different such that the numbers of samples are consecutive prime numbers. Furthermore, when the received signals at CRs are faded or shadowed, the performance of MS is analytically evaluated. Numerical results show that the proposed system can significantly enhance the wideband spectrum sensing performance while requiring low computational and implementation complexities.