Low-rank Matrix Completion based Malicious User Detection in Cooperative Spectrum Sensing

In a cognitive radio (CR) system, cooperative spectrum sensing (CSS) is the key to improving sensing performance in deep fading channels. In CSS networks, signals received at the secondary users (SUs) are sent to a fusion center to make a final decision of the spectrum occupancy. In this process, the presence of malicious users sending false sensing samples can severely degrade the performance of the CSS network. In this paper, with the compressive sensing (CS) technique being implemented at each SU, we build a CSS network with double sparsity property. A new malicious user detection scheme is proposed by utilizing the adaptive outlier pursuit (AOP) based low-rank matrix completion in the CSS network. In the proposed scheme, the malicious users are removed in the process of signal recovery at the fusion center. The numerical analysis of the proposed scheme is carried out and compared with an existing malicious user detection algorithm.