Intelligent clustering for balanced energy consumption in wireless sensor networks

In this paper, we present a new intelligent clustering protocol based on self -Organizing Map neural network in order to balance the energy consumption in Wireless Sensor Networks. The clustering would be done based on two important criteria, energy level and coordinates of sensor nodes. Energy based clustering can cause more balanced energy consumption compared to topologic clustering alone, through creating clusters with equal energy levels. As a result, this new clustering design can prevent from early death of the nodes and allow for random death of them. Moreover, a new cost function has been presented to decide about selecting an appropriate cluster head for each formed cluster based on multiple criteria in which the weights of these criteria are determined through an AHP method. Simulation results proved better performance of proposed clustering design and cluster head cost function in extending network lifetime compared to two previous protocols.