Energy-saving clustering algorithm based on improved PSO for WSNs

The LEACH algorithm selects the Cluster Head (CH) randomly with the same probability; however, it cannot guarantee that the number of clusters in each round is the best. In addition, non-Cluster Head node (non-CH) in the process of selecting the cluster does not take the global energy consumption into account. In this paper, considering the disadvantages of cluster's structure in LEACH, we adopt the Particle Swarm Optimization (PSO) algorithm for choosing CHs. Based on the adjustment of the inertia weight and the acceleration factor, we amend the disadvantages of PSO in terms of convergence speed in the late stage, which make the particle fall into local optimum very easily. An evaluation function is proposed to calculate which cluster the non-CH node will belong to. Simulation experiments show that the proposed algorithm is more superior than LEACH in clustering efficiency, and the lifetime of the network is significantly extended.