Energy-efficient clustering algorithm based on energy variance for Wireless Sensor Networks

In Wireless Sensor Network, the total energy consumption determines the lifetime of the entire network. As a result, a number of recently-designed energy-efficient routing algorithm stated that clustering approach has an important issue for organizing a network into a connected hierarchy and increasing the network lifetime. The clustering hierarchical model consists in two main steps, cluster heads selection and assignment of each node to one cluster. This paper proposes an enhancement of Low Energy Adaptive Clustering with deterministic cluster-head selection by introducing a new parameter for electing cluster-head. Nodes having the highest remaining energy and the lowest energy variance consumption become cluster heads with high probability. The additional variance parameter takes into account energy consumption dispersion if the considered node is elected as cluster-head. This dispersion highly depends on the relative positioning of the node to the base station. Variance keeps the energy consumption of the node in the preceding rounds. This is useful for predicting node status when elected as cluster-head in the current round in order to balance the network energy. Simulation results show that our proposal extends the networks lifetime compared to recent cluster-based protocol. More-over, the energy consumption is balanced between the wireless sensor nodes.