Threshold optimisation with Bayesian approach in covariance absolute value detection method

This study investigate statistical properties of Bayes cost, it proposes new group-based detection models for Cognitive Radios (CR) systems using this statistic. In commonly used spectrum sensing applications, multi-user or multi-antenna detection scenario with different optimisation methods are recommended to increase the detection under high SNRs. In this study, the optimum threshold for Covariance Absolute Value detection methods is determined using the Bayesian Cost. Thus, the detection scenario is given for the proposed method and the results are compared with the traditional Covariance Absolute Value method. Simulations based on random communication signals are presented to verify the methods. © 2021 Informa UK Limited, trading as Taylor & Francis Group.