Robust Statistics Evidence Based Secure Cooperative Spectrum Sensing for Cognitive Radio Networks

Abstract

Cognitive radio networks (CRNs), an assemble of smart schemes intended for permitting secondary users (SUs) to opportunistically access spectral bands vacant by primary user (PU), has been deliberated as a solution to improve spectrum utilization. Cooperative spectrum sensing (CSS) is a vital technology of CRN systems used to enhance the PU detection performance by exploiting SUs' spatial diversity, however CSS leads to spectrum sensing data falsification (SSDF), a new security threat in CR system. The SSDF by malicious users can lead to a decrease in CSS performance. In this work, we propose a CSS scheme in which the presence and absence hypotheses distribution of PU signal is estimated based on past sensing received energy data incorporating robust statistics, and the data fusion are performed according to an evidence based approach. Simulation results show that the proposed scheme can achieve a significant malicious user reduction due to the abnormality of the distribution of malicious users compared with that of other legitimate users. Furthermore, the performance of our data fusion scheme is improved by supplemented nodes' credibility weight.