Equalizing cluster lifetime for sensor networks with directional data traffic to improve energy efficiency

Abstract

In order to balance the burdens among clusters in a sensor network, traditional clustering algorithms proposed for energy saving usually organize clusters to make them similar in size. However, the directional and uneven data traffic towards the data sink places a greater burden on the clusters near the sink, which makes them drain energy more rapidly than other clusters in homogeneous sensor networks. The areas covered by these clusters are not monitored once they have died prematurely. This paper proposes an algorithm to equalize cluster lifetimes throughout the network by organizing the clusters in such a way that the cluster energy store is proportional to its power consumption, which is relative to the traffic burden. Simulation results show that this algorithm does indeed balance power consumption throughout the network. It increases both the first-node lifetime (when any node dies) and the connectivity lifetime (when the connectivity between the sink and the network is lost) of sensor networks. Furthermore, it improves network performance by preventing some areas from losing coverage due to premature node death before the connectivity lifetime.