A Convergent Algorithm for Energy-Balanced Cluster-Heads Selection in Wireless Sensor Networks

Abstract

Due to the limited energy of sensor nodes, it is a research goal that the lifetime of sensor networks is prolonged by transmitting the sensed data to the base station in an energy-saving way. Previous algorithms aim at reducing the average energy consumption rate to extend the network lifetime. However, some nodes sometimes may be served as the cluster-head too many times to conserve their energy, resulting in reduced network lifetime. Thus, the large deviation of network lifetime makes these algorithms impractical. This paper proposes a new clustering algorithm which not only reduces the average energy consumption rate, but also converges the residual energies of all nodes on a small interval. Based on the two-region cluster-heads selection mechanism, the coordinator adaptively adjusts the far-near regions to converge the energies of all nodes on a small interval. With the exclusion-circle of cluster-heads, cluster-heads can be distributed evenly in a spatial respect for each round, resulting in reduced energy consumption. The simulation results show that the proposed algorithm not only makes cluster-heads distribute evenly in a spatial respect but also converges the residual energies of all nodes on a small interval, resulting in extending the network lifetime significantly and stably.