Energy-Efficient Link-aware PSO-Based Clustering Algorithm in Wireless Sensor Networks (WSNs)

Recently Wireless Sensor Networks (WSNs) has turned into a popular matter of research area, because of its flexibility and dynamic nature. It is proven that the clustering technique, as a multi objective optimization, is the most effective solution to have minimum energy consumption. The goal of the clustering technique is to divide network sensors into clusters each of which has a cluster-head (CH) responsible to collect, aggregate and send sensed data to the base station (BS). The recent researches shown that such multi objective optimization in WSNs can be solved through well adapted an evolutionary algorithm. In this paper an improved k-mean clustering model powered by Particle Swarm Optimization (PSO) algorithm is presented. This is called Link-aware PSO, LSPO. The proposed model utilizes two-phase optimization by applying different fitness function. At the first phase, it selects Primary Cluster-heads based on improved Intra-Cluster Distance metric as fitness function in PSO algorithm to give primary CHs. In the second phase each primary CHs selected are evaluated by link quality and energy metrics to select the best ones as final CHs. Simulation results showed that the proposed algorithm outperforms LEACH and PSO-C algorithms in term of performance, prolonging network lifetime and energy saving.