Energy Efficient k-Means Clustering Algorithm for WSN using Virtual Grid and Sink Mobility

Hierarchical based routing is a routing technique that create virtual hierarchy among network sensor nodes. The network is split up into small clusters that effectively increase network lifetime. The traditional clustering algorithms available in literature have some limitations. These algorithms do not consider the self-organized dynamic topology that is present in WSNs. Uneven distribution of cluster heads resulted due to these limitations. This uneven distribution of cluster heads effected the overall network energy consumption. In our work we proposed a virtual grid and sink mobility-based routing scheme with k-means clustering technique for a WSN. It combines the effect of dynamic route adjustment due to sink mobility with k-means technique for cluster head election of each grid cell or cluster. This scheme divides the network into small grid cells and then select cluster head for each grid using k-means algorithm. After initial cluster head selection, a virtual backbone network established and mobile sink moves at a relatively optimal predefined path inside the network. Dynamic routes re-adjustment cost incurred by sink mobility is minimized by virtual backbone network. Proposed scheme enhances the stability of the network and increase the overall lifetime of the sensor network.