A PCA Based K-Means Clustering Algorithm for Wireless Sensor Nodes

Divleen Kaur; Ravi Kumar

doi:10.37591/jowet.v2i2.437

Abstract

This paper presents a novel and improved approach for K-Means Clustering of wireless sensor networks, by using Principal Component Analysis for data reduction on the raw data. A wireless sensor network consisting of 100 nodes is classified into three different clusters using PCA based K-Means Algorithm. Davies-Bouldin Index is used as a parameter to check the effectiveness of the clustering algorithm. Experimental results demonstrate that the PCA based K-Means Algorithm increases the quality of clustering and assigns sensor nodes to their appropriate cluster more efficiently and hence provides a more robust and effective clustering of the wireless sensor network as compared to the conventional K-Means Algorithm.Keywords: PCA, K-means, davies-bouldin index, wireless sensor networks Cite this ArticleDivleen Kaur, Ravi Kumar, A PCA Based K-Means Clustering Algorithm for Wireless Sensor Nodes. Journal of Web Engineering & Technology. 2015. 2(2): 6–10p.