An Efficient Clustering Algorithm Using Evolutionary HMM in Wireless Sensor Networks

Energy efficiency should be considered as a key design objective in wireless sensor networks (WSNs), since a sensor node can only be equipped with a limited energy supply. Clustering is one of the well-known design methods for managing the energy consumption in WSNs. Rotating role of cluster heads (CH) among nodes in these networks is an important issue in some of clustering methods. Directly collecting information about the energy level of nodes in each round increases the cost of CH role rotation, in the field of centralized hierarchical methods. In this paper, we proposed a centralized clustering algorithm that utilize hidden Markov model (HMM) optimized by particle swarm optimization (PSO) to predict the energy level of the network. In the next step, the appropriate CHs are selected by PSO algorithm. Our proposed method reduces the cost of clustering and in the mean time increases clustering performance. Evaluation results demonstrate by comparison with famous clustering algorithms, our scheme is energy efficient and increase network life time.