A novel K-means L-layer algorithm for uneven clustering in WSN

Clustering of nodes in a WSN is one of the proven ways to achieve increased lifetime of the network. Many novel algorithms continue to be proposed to achieve this objective. A survey of the literature also suggest that uneven clustering with less nodes closer to the base station achieves greater efficiency than same number of nodes in all clusters. This is due to larger overheads for the nodes closer to the base station. This work proposes a novel K-Means L Layer algorithms which leads to the creation of clusters with lesser number of nodes closer to the base station as opposed to the ones far away from it for randomly deployed nodes. The proposed algorithm is a modification of the K Means algorithm which provides even clustering. Further another contribution of this paper is the study of energy consumption of the nodes with regards to the data packet optimization.