Optimal Decision Fusion Strategy in Cognitive Radio Networks

Abstract

In this paper, we analyze performance of cooperative spectrum sensing under counting rules when exponential model involving the spectrum usage (SU) is utilized to characterize the burst nature of primary user (PU) link. Considering both the reliability and the efficiency performance of sensing, we use the average error probability (AEP) in final decision making as metric. We derive a closed-form expression of AEP by classifying cognitive transmission into six cases. In order to minimize the AEP, we consider the optimization of cooperative sensing with energy detection, and find the optimal fusion rule among all of these counting rules. Finally, we verify the theoretical analysis by numerical results.