Tri-threshold cooperative spectrum detection for cognitive radio based on weighing

In cognitive radio (CR), energy detection is a simple and feasible method to get reliable detection results on the presence of primary user (PU) in perfect channel. In imperfect channel, single-threshold cooperative spectrum detection has been proposed to improve energy detection performance, however it is sensitive to noise interference. In order to solve this problem, double-threshold cooperative spectrum detection has been proposed, however, works not well when secondary user (SU) cannot distinguish signal from noise. In this paper, tri-threshold cooperative spectrum sensing based on weighing is proposed to improve the detection performance when the SUs´ SNR is low. In the proposed scheme, the energy statistic is divided into four subareas and the two middle confusion areas are processed according to the average SNR of SUs locating at the confusion areas. The simulation shows that compared with the traditional methods, the tri-threshold cooperative detection can improve the detection performance especially when SUs cannot distinguish signal from noise.