Trust Model-based Secure Cooperative Sensing Techniques for Cognitive Radio Networks

Cooperative spectrum sensing has been shown to enable Cognitive Radio (CR) networks to reliably detect licensed users and avoid causing interference to licensed communications. However, the performance of the scheme can be severely degraded due to presence of malicious users sending false sensing data. In this paper, we propose trust model-based cooperative sensing techniques to reduce the harmful effect of malicious users in cooperative sensing process. First of all, analysis model of anomalous behavior is devised to identify the malicious users. Then we employ PID (Proportional-Integral-Derivative) like controller to calculate the credit value of nodes, which is used as the weight of WBD (Weighted Bayesian Detection) to make spectrum decision. Simulation results demonstrate that, comparing with the existing methods, the proposed scheme performs better especially in the case that there exist a large number of malicious nodes.