A spectrum allocation method based on random drift swarm optimization algorithm

Efficiently of spectrum allocation is the key for dynamic spectrum access in the Cognitive Radio Network (CRNs). Based on graph coloring model, this paper proposes a random drift particle swarm optimization (RDPSO) algorithm to allocate the spectrum, with the purpose of improving the reward and fairness of network. The RDPSO algorithm introduces the thermal coefficient and drift coefficient, and it shows the faster convergence and better performance optimization than particle swarm optimization (PSO). The simulation results shown that this algorithm is flexible, which has high location accuracy and fast convergence speed. And compared to other three algorithms, RDPSO algorithm improved reward and fairness of network and convergence.