Support Recovery for Multiband Spectrum Sensing Based on Modulated Wideband Converter with SwSOMP Algorithm

Abstract

The Modulated Wideband Converter (MWC) can provide a sub-Nyquist sampling approach to sense sparse multiband analog signals and reconstruct the frequency support set. However, the existing SOMP reconstruction algorithms need a priori information of signal sparsity. This paper applies the SwOMP algorithm to the CTF (Continuous-To-Finite) block of MWC. The SwSOMP algorithm uses stage-wise weak selection in SOMP, and it can reduce computational cost and solve large scale problems. It does not need prior information of signal sparsity, and the frequency support can be reconstructed blindly. The simulation results demonstrate that, MWC system with SwSOMP algorithm, compared with the SOMP algorithm, can use less number of channels, achieve higher percentage of correct support recovery blindly, and reduce the sampling rate of the system.