An Energy-Efficient Decentralized Clustering Protocol for Wireless Sensor Networks

This paper presents a novel clustering protocol, the Decentralized Energy Efficient cluster Propagation (DEEP) protocol, that manages the communication of data while minimizing energy consumption across sensor networks. The paper also presents an Inter-Cluster Routing protocol (ICR) that is compatible with the proposed clustering technique. The proposed clustering protocol uses the unique technique of forming clusters using a pre-selected initial cluster head that initiates the advertisement process. The initial cluster head also identifies its cluster members as well as new cluster head candidates based on the relative distance between sensor nodes and their remaining energy levels. Because this model results in a balanced load among cluster heads, protocol overhead due to frequent re-clustering is eliminated. Simulation results demonstrate that the DEEP protocol distributes energy consumption approximately eight times better than the LEACH clustering scheme. In addition, the DEEP protocol substantially reduces total data communication and route setup energy consumption in the network compared to the LEACH protocol.