Probability‐based fusion rule in cooperative spectrum sensing with impact of geographic location

Abstract

This paper investigates the impact of geographic locations of secondary users (SUs) on cooperative spectrum sensing (CSS) in cognitive radio networks (CRNs), where SUs send their local sensing results to a fusion center (FC) and the FC makes a global spectrum decision by a specific fusion rule. The purpose of this paper is to propose a fusion rule that takes into account the impact of geographic locations on CSS, and identify the optimal decision threshold of the proposed fusion rule in order to maximize the average achievable throughput of the secondary network subject to a collision constraint. To quantitatively capture the impact of geographic locations on CSS, a metric called absolute impact is defined, which is determined by the probability of detection and the probability of false alarm. Meanwhile, another metric is defined called relative impact to characterize the impact of SU's incorrect local sensing result on CSS. Based on the defined metrics, a probability-based fusion rule is proposed, and the core idea is that the global spectrum decision of CSS is jointly determined by the local sensing results and the absolute impacts of SUs. Then the average achievable throughput of the secondary network and the collision probability are derived, both of which are proved to decrease with the decision threshold of the proposed fusion rule. By tackling the trade-off between the accuracy of spectrum sensing and spectrum access opportunity caused by the decision threshold, an approximate algorithm based on the idea of the bisection method is proposed to obtain the optimal decision threshold that maximizes the average achievable throughput of the secondary network while protecting primary transmission to a predefined extent. Numerical results are carried out to verify the theoretical analyses and show the flexibility of the proposed fusion rule.