Reconstruction failure detection for wideband spectrum sensing with modulated wideband converter based sub-Nyquist sampling

Modulated wideband converter (MWC) based sub-Nyquist sampling can be used for wideband spectrum sensing in cognitive radio networks. In this paper, a method is proposed for detecting whether the reconstruction process is unsuccessful due to non-sparse spectrum in MWC based wideband spectrum sensing to avoid causing harmful interference to primary users. The proposed method exploits the correlation between two consecutively recovered support sets. Simulation results show that the method can judge whether the reconstruction is successful or not with high probability. Compared with traditional spectrum sensing without reconstruction failure detection, the proposed method can lower the interference probability to primary users greatly when spectrum is not sparse.