A Many-objective Chaotic Whale Optimization based Automatic Clustering Algorithm for Distributed Data Analysis in WSN

Abstract

The potential applications of Wireless senor networks (WSN) has increased rapidly in weather reporting, target tracking, oxygen content monitoring and many other applications due to their low cost and less power requirements. The sensor nodes in WSN collect the dataset which is transferred to base station for further processing. Clustering is a key part of processing WSN dataset. To reduce the burden of base station researchers are using distributed approaches for clustering where individual sensor node performs cluster analysis with collaboration of neighboring nodes. Nature inspired algorithms provides better clustering results for WSN dataset in single and multi-objective framework. As the complexity of WSN data set increases the single and multi objective algorithms fails to give correct partitions when applied for clustering. To alleviate this problem this paper introduces an automatic distributed many-objective chaotic whale optimization algorithm (ADCWOA) to perform distributed clustering in WSN datasets. The proposed ADCWOA outperforms automatic distributed many objective particle swarm optimization (ADPSO) and automatic distributed K-Means (ADK-Means).