Dynamic attack probability based Spectrum Sensing against Byzantine attack in Cognitive Radio

Spectrum sensing plays an important role in improving the spectrum utilization in cognitive radio. However, the existence of Byzantine attackers impairs the overall performance of Cognitive Radio Networks (CRNs) severely, no matter in detection accuracy, throughput or energy efficiency and so on. Therefore, it is vital to identify attackers and decrease their abominable influences. In this paper, a novel attacker-identification algorithm with dynamic attack probability is proposed, which can enhance the overall detection probability markedly and effectively reduce the false alarm probability. Meanwhile, a weight coefficient is explored to energy detection (ED) to promote the robustness of the system. Simulation results shows promising performance of proposed algorithm compared to previous works.