Optimization for Weighed Cooperative Spectrum Sensing in Cognitive Radio Network

In this paper, an optimal weighed cooperative spectrum sensing strategy based on data fusion is investigated in cognitive radio network. Cognitive radios sense the channels by energy detection independently and send their results to a fusion center, in which the observed data are fused by the specific weighing. The optimal sensing problem, which seeks to minimize interference and maximize throughput by keeping the probabilities of false alarm and detection within the allowable limit, is formulated. In particular, both the cooperative detections in single channel and multi-channels are analyzed, and the optimal weighed factors are obtained by Cauchy-inequality. Based on the weighing, we transform the non-convex optimal problem of multi-channel sensing with double parameters and nonlinear constraints into a convex problem with single parameter and linear constraints, which can be easily solved. The simulation shows that the proposed algorithm can achieve lower interference and higher throughput with less computing complexity, and the detected performance of each sub-channel can also be guaranteed. It also indicates that there is a conflict between improving throughput and decreasing interference, and the proposed algorithm can make better use of spectrum by balancing the conflict.