K-SEP: A more stable SEP using K-Means Clustering and Probabilistic Transmission in WSN

Abstract

A Wireless sensor network consists of large number of sensor nodes having different energy levels deployed for sensing and monitoring task. As energy is the major constraint in wireless sensor network these nodes are mostly organized into clusters to save energy. In this paper a modified approach for Stability Election Protocol called K-SEP is proposed which aimed at improving the stability period by balancing the energy consumption of the nodes. The proposed protocol uses K-means algorithm for uniform clustering and select cluster heads based on current energy levels and distance. Also transmission from cluster members to cluster head is made probabilistic. Simulation result shows that proposed protocol have better stability period than SEP.