Distributed cooperative spectrum sensing with multiple coalitions and non-ideal reporting channel

Abstract

Spectrum Sensing (SS) is a crucial task of Cognitive Radio (CR) to avoid the destructive interference of the unlicensed users to the licensed users of the spectrum and to find the accessible spectrum for improving the spectrum utilization. But detection of licensed users is often compromised due to issues such as multipath fading, shadowing, and receiver uncertainty. So to deal with these issues, Cooperative Spectrum Sensing (CSS) has been proposed. There are two classes in CSS: centralized and decentralized or distributed. In this paper, we compare a fully distributed CSS technique using Consensus Algorithms with the existing OR fusion rule based CSS scheme under AWGN, Rayleigh and correlated log-normal shadowing environments. We also compare existing OR rule with a distributed coalition sensing scheme and study the effect of error in reporting channel. Simulation results show that distributed consensus scheme outperforms OR fusion rule, thus showing the importance of distributed schemes. But in the case of distributed coalition sensing and OR fusion rule, both show somewhat same performance. We found that, with increasing probability of error in reporting channel, the probability of false alarm increases drastically showing a degradation in detection performance.