Robust Cooperative Primary User Detection in Malicious Cognitive Radio Networks

Cognitive radio networks (CRN s) rely heavily on dependable detection of the presence or absence of primary users (PUs). Cooperative secondary users (SUs) forwarding the received PU signals to a fusion center (FC) offer diversity to improve PU detection performance by a centralized approach. Several schemes have been proposed to identify malicious SUs, which for different reasons might be willing to falsify the signal forwarded to the FC. Existing proposals for FC based PU detection are insufficient concerning the achievable performance for an arbitrary number of malicious SUs and/or privacy issues. The sequential estimation-detection (SED) scheme presented here is simple and robust for different system scenarios and arbitrary numbers of malicious SUs. The latter cannot be identified so that privacy is guaranteed. A full system description and simulations are provided to characterize the scheme's performance.