A novel centralized clustering approach based on K-means algorithm for wireless sensor network

This paper describes wireless sensor networks for a novel hierarchical routing protocol based on the centralized clustering approach inspired on the k-means method in order to organize sensor networks into clusters and get a better performance parameter. This contribution tries to enhance the clustering process in LEACH-C protocol by applying k-means algorithm in order to produce a new cluster scheme and therefore lengthen the lifetime of the sensor network. As far as the simulation is concerned, our approach was implemented in NS2 simulator and we have evaluated the following parameters of performance: Average end-to-end Delay, Packet Delivery Ratio, Average Energy Consumption, Average Throughput and Control Routing Overhead. Compared with LEACH-Centralized protocol, the simulation results demonstrate that LEACH-Ckmeans algorithm can effectively hugely reduce the overhead and it gives no loss of packets to the destination which leads to increased network performance.