Spatial Correlation and Centroid Based Clustering in Wireless Sensor Network

Abstract

Listen

Translate article

Wireless Sensor Networks (WSN) is a dense deployment of sensor nodes that constantly monitoring physical phenomenon's. Due to high density, typically $20 nodes/m^{2}$ , nearby located sensor nodes often sense similar reading i.e. they are spatially correlated. Hence, they are transmitting redundant data often causing unnecessary energy consumption in battery operated sensor node. However, grouping sensor nodes together will reduce energy consumption. Traditionally, clustering is used to accomplished this task. In this paper, we propose a clustering algorithm using spatial correlation which is different from other clustering algorithms. Proposed algorithm groups sensor nodes with similar readings into one cluster, such that, it is enough to report a single reading from the entire group, which will reduce energy consumption and increase network lifetime. Between the group of nodes, one node is selected as a cluster head using centroid method. The sensor node which has minimum distance from cluster centroid point is chosen as a cluster head. Hence, clustering using spatial correlation can minimize the residual energy, maximize the network performance, increases network lifetime and decreases network traffic.