A low complexity cluster based data fusion to defense against SSDF attack in cognitive radio networks

In cognitive radio networks, the security attacks disrupt the spectrum sensing process. One of the most threatening attacks is spectrum sensing data falsification (SSDF) attack, where some of the network users are malicious users and by sending their incorrect observations degrade the decision making process. In this paper, to counter this kind of attack, we propose a method based on clustering the secondary users and estimating some unknown parameters of their received power. In the proposed scheme, we consider two different scenarios for the attack strategies: deterministic attack and WGN (White Gaussian Noise) attack. Due to the use of clustering and the low computational complexity of the proposed method, the complexity order reduces significantly. Moreover, our scheme has good performance in detecting the malicious users of each cluster.