A Dynamic Energy-Efficient Chain Formation Scheme for PEGASIS in Wireless Sensor Networks

Data aggregation is a promising way to reduce the redundant data communication in wireless sensor networks and hence improves the network lifetime. Among the various data aggregation protocols, it is reported that PEGASIS is better than LEACH and HEED. PEGASIS is a chain based aggregation protocol, where each node communicates only with its closest neighbour and on its turn to become a leader, transmits aggregated data to the base station. An energy-efficient variation of PEGASIS, called concentric clustering scheme divides the network into concentric clusters. The leader node is the extreme-end node of a cluster who is responsible for gathering and aggregating sensed data. However, in the concentric clustering scheme, the leader selection mechanism does not ensure the connectivity between the leaders of each cluster. Moreover, it does not consider energy level of the nodes during leader selection. In this paper, we propose an improvement to concentric clustering scheme by designing a mechanism that ensures the connectivity among the leaders and considers the residual energy level of a node during leader selection process. The simulation result shows that the proposed scheme has significant improvement in terms of network lifetime, connectivity, and number of alive nodes.