Defending Cooperative Spectrum Sensing From Byzantine Attacks: An Effective Entropy-Based Weighted Algorithm

Cooperative Spectrum Sensing (CSS) is an effective approach used to determine accessible spectrum in cognitive radio networks (CRN). However, CSS is susceptible to being attacked by malicious users (MUs) launching the Byzantine attack, which degrades the CSS performance. This study focuses on reducing the effects of MUs on CSS performance and identifying the MUs in the CRN. We propose an effective entropy-based weighted algorithm for CSS, which not only enhances CSS performance but also enables the detection of MUs in the CRN. Extensive analysis demonstrates the superiority of the proposed algorithm. Importantly, the algorithm is suitable for both homogeneous and heterogeneous scenarios, accommodating CRNs with various levels of similarity or dissimilarity in the local sensing capabilities of the cooperating secondary users (CSUs).