Effect of Number of Users and Number of Clusters Using Distributed Cooperative Spectrum Sensing Over Hoyt Fading Channel

Abstract

With the growth of mobile subscribers, the need for wireless devices is increasing. Present radio spectrum framework limits this growth due to spectrum scarcity. Cognitive radio is an intelligent wireless system, and spectrum sensing is an important feature of cognitive radio network (CRN). Sensing vacant spectrum becomes difficult due to problems such as multipath fading, shadowing and hidden terminal in non-cooperative spectrum sensing technique. Cooperative spectrum sensing (CSS) is used to overcome these problems. CSS is of two types: centralized and distributed. In this paper, detection performance of cluster-based distributed CSS is investigated in Hoyt fading environment. Secondary users (SUs) are divided into clusters, and under distributed CSS scheme different fusion rules are applied on them. Number of users and number of clusters are changed, and their effect on performance is investigated. It is observed that at 0.001 false alarm probability (Pf), if we increase the number of users from 4 to 9, the probability of missed detection (Pm) decreases by 58.51% and Pm decreases by 78.59% when the number of users increases from 4 to 25. Probability of detection (Pd) is complementary of probability of missed detection (Pm).