On the Detection Performance of Cooperative Spectrum Sensing Using Particle Swarm Optimization Algorithms

In this paper, the main aim is to optimize the probability of detection in cognitive radio (CR) networks by using different variants of particle swarm optimization (PSO). Finding the optimal weighting coefficient vector in soft-decision fusion based cooperative spectrum sensing is a challenging task that is crucially needed to improve the detection performance. The performance of standard PSO (SPSO) and five other PSO variants named as self-organizing hierarchical PSO with time acceleration coefficients (HPSO-TVAC), median-oriented PSO (MPSO), centripetal accelerated PSO (CAPSO), gravitational particle swarm (GPS), and cooperative gravity based PSO (CGPSO) is analyzed while searching for the optimal weighting coefficient vector at which the overall probability of detection is maximized, given a fixed probability of false alarm. The best achievable fitness, convergence speed, and stability of the PSO variants are compared. The simulation results show that GPS is the best choice among all other PSO variants to improve the detection performance in the given CR network deployment.