Cooperative Spectrum Sensing using Rule based Hard Decision and Soft Decision with Bayesian Optimized Support Vector Machine

Abstract

Cognitive radio technology has emerged as an operational substitute for increasing the number of users of broadband data services in various frequency bands. This article presents the results of an analysis of the effectivenessof the fusion rule for joint spectrum detection in cooperative spectrum sensing (CSS). The energy detectors-basedfeature vector is considered for ML training purpose. The paper proposes support vector machine (SVM) basedmodelling for training and testing in CR. Further Bayesian optimized SVM is proposed to claim higher detection rate.The proposed method yields 0.84 detection rate at 0.1 probability of false alarm