Soft Reputation-Based Secure Cooperative Spectrum Sensing

Cognitive Radios (CRs) sense the radio spectrum environment dynamically and automatically in order to find underutilized spectrum and then exploit it intelligently. In face of the challenging propagation environments characterized by shadowing and fading effects, cooperative spectrum sensing (CSS) has been proposed to perform reliable spectrum sensing by taking advantage of the spatial diversity in multi-user Cognitive radio networks. In CSS, each secondary user sends its sensing data to the combining node, which makes a decision on the presence of the primary signal. However, CSS are typically very vulnerable to attacks where malicious CRs falsify sensing data. To identify misbehaved CRs and alleviate their adverse effects on CSS performance, a secure CSS scheme based on soft reputation has been proposed in this letter. Contrary to existing reputation-based secure CSS schemes, where reputation of CRs is assigned artificially, reputation of CRs is calculated utilizing the statistics of primary band occupancy. Simulation demonstrates the effectiveness of the proposed scheme.