Reputation-based Beta reputation system against SSDF attack in cognitive radio networks

In cognitive radio networks (CRNs), cooperative spectrum sensing performed by multiple users is able to improve the sensing performance of the cognitive system since available spectrum holes can be detected and utilized, thus increasing spectrum utilization efficiency. However, cognitive users are vulnerable to spectrum sensing data falsification (SSDF) attacks, which arises as one of the most serious security threats in CRNs. In the SSDF attack, attackers send erroneous local reports to the fusion center during cooperative spectrum sensing process to mislead the fusion center to make a false global decision. The security of the network guarantees that wireless communication can be carried out smoothly. Therefore, the research on secure spectrum sensing is necessary for resolving the problem of SSDF attacks and ensuring the security of the whole system. In this article, a dynamic trust mechanism based on Beta reputation system is proposed to resist SSDF attacks. We introduce the time proportion coefficient, which is set in line with the historical sensing behaviors of cognitive users, and the reputation value of users can be obtained accordingly, besides, the rise factor and fall factor are introduced as well. Cognitive users that are qualified to participate in the cooperative sensing can be selected in accordance with users' reputation values. The final global decision can be obtained by linear fusion method in the center data fusion. Simulation results show that, in the proposed reputation mechanism, the SSDF attack is effectively suppressed since the reputation of malicious cognitive users are significantly reduced, thus improving the accuracy of cooperative spectrum sensing.