Analyzing the Performance of Centralized Clustering Techniques for Realistic Wireless Sensor Network Topologies

Abstract

Clustering techniques in wireless sensor networks enables energy efficient coordination among the densely deployed nodes for data delivery till the base station. Many clustering protocols have been suggested in the recent past. The topology of the nodes, mostly seen in the literature, is of random type. This paper analyzes the performance aspects of various centralized clustering techniques for wireless sensor networks. LEACH-Centralized, KMeans-CP, FCM-CP and HSA-CP protocols have been compared with respect to clustering and data delivery process for various realistic topologies. The simulations were performed for these protocols and performance of the protocols has been critically analyzed. HSA-CP clustering method performs better compared to other techniques for almost each topology examined in the paper.