An Efficient Dynamic Energy Aware Neighbor Oriented Clustering

Abstract

Network clustering consists of partitioning a large-scale network into clusters. In each cluster, we can identify a simple node and a particular node named cluster-head (CH). All simples nodes transmit their data to the cluster-head who aggregate them and send the packet to the sink. This technique reduces the amount of information and energy consumption. CHs are characterized by their high energy consumptions compared to other nodes. Thus, CH should be appropriately selected to preserve the overall network energy and then improve the lifetime. For this reason, we propose in this paper, a new routing protocol based on a hybrid metric called Dynamic Energy Aware Neighbor Oriented Clustering (DEANOC). This protocol elects a CH in each cluster of a network. The node with the highest residual energy, the closest to the station and less neighbor is chosen as CH. If the amount of energy of CHs is less than the threshold (TEN), the clustering algorithm is executed. Simulation results show that DEANOC improves the network lifetime and reduces the energy consumption compared to a similar protocol.