Improved weighted cooperative sensing algorithm based on distributed optimization in cognitive radio networks

Focusing on the contradiction between spectrum sensing performance and resource utilization in cognitive radio networks, an improved weighted cooperative sensing algorithm based on distributed optimization is proposed. The algorithm can enhance the spectrum sensing performance and reduce both network overhead and sensing time. Firstly, double-threshold energy sensing is determined according to the maximal probability of false alarms and missing detections. By comparison of the detected signal energy and double threshold, the cognitive radio users are separated into the trusted group and the incompletely trusted group. Secondly, an utility function of cognitive radio networks is defined. To achieve the dynamic adjustment for an energy sensing threshold, a subgradient algorithm is employed to optimize the utility function distributedly and cooperatively. The selection of cognitive radio users from the incompletely trusted group is accomplished according to the rate of convergence of optimization. Finally, an overall decision is obtained at the fusion center. The efficiency of the algorithm is tested by simulation in dynamic cognitive radio networks.