Thresholds selection for censored energy detection based two witnesses rule for cognitive radio networks

Abstract

In this paper, cooperative spectrum sensing based censored energy detection is considered in order to improve the accuracy of the decision made about the spectrum availability. Proposed two witnesses fusion rule is used to derive a spectrum sensing final decision. However, the accuracy of this decision cannot be ensured unless the best choice of the two decision thresholds is made such that the probabilities of mis-detection and false alarm are minimized. In this context, one primary cell and one secondary cell are considered and closed forms of the probabilities of detection, false alarm, occupancy and mis-detection are defined. Based on these probabilities' closed forms, a thresholds selection scheme is presented. Simulation results prove the coherence between the analytical expressions of the earlier mentioned probabilities and the simulated ones when selected thresholds are used.