A Trust-Based Cooperative Spectrum Sensing Scheme against SSDF Attack in CRNs

Cooperative spectrum sensing (CSS) can effectively improve sensing accuracy and suppress channel fading effects. But it can be threatened by malicious users (MUs). MUs may try to forge sensing results to indicate that the primary user exists even when there is no primary user (PU). So MUs can monopolize the spectrum usage, thereby depriving other users of their spectrum opportunities. Based on a more general CSS which considers both unintentional misbehaved users and intentional MUs in cognitive radio networks (CRNs), we propose a trust-based CSS scheme. To avoid invalid data transmission and reduce the consumption of bandwidth, we select k secondary users (SUs) to perform CSS based on their SNR instead of all-member participation. Furthermore, we aim at preventing unintentional misbehaved users from being mistaken for intentional MUs and excluded from the cooperation. The decision fusion center (DFC) combines local decisions based on the trust value of cooperative users. This paper tries to alleviate the passive impacts of MUs instead of detecting and permanently eliminating them from the candidates of CSS. In the proposed scheme, we find the optimal detection threshold to meet a desired sensing accuracy. Numerical results show that the proposed trust-based cooperative spectrum sensing scheme can avoid significant reduction in performance of CSS with SSDF attack.