Optimization of multiband cooperative spectrum sensing with modified artificial bee colony algorithm

Abstract

In this paper, the multaiband cooperative spectrum sensing problem in cognitive radio network is investigated, which focuses on maximizing the total throughput under the constraint of the interference to primary users by jointly optimizing weight coefficients and decision threshold. Due to the non-convex characteristics of this optimization problem, a modified artificial bee colony (MABC) algorithm is proposed to deal with this problem, in which some improved mechanisms, such as mutation and crossover factors are introduced in ABC to enhance the diversity and improve the searching ability. Classical benchmark functions are employed to evaluate the searching ability of MABC when compared with PSO, GA, ABC and EA-ABC. Simulation results have been provided to validate the promising performance of MABC over the other intelligent evolutionary algorithms when applied to both the classical benchmark functions and multiband cooperative spectrum sensing scenario.