A Realistic Weighted Clustering algorithm for data gathering in single hop cell phone based sensor network

Abstract

We present Weight based Realistic Clustering algorithm (WRCA) for energy efficient data collection from human based wireless network such as cell phone based sensor network. The WRCA creates better load balanced and stable clusters for data gathering in a realistic human mobility scenario as compared to WCA algorithm of mobile adhoc networks. Traditional way of stable clustering used to select stable Cluster Heads (CHs) based on instantaneous mobility criteria. While our algorithm incorporate only a pause time weighting parameter to leverage pause time distribution of human mobility for selection of more stable CHs. Load balancing aspect of WRCA tackles realistic scenario of inhomogeneous node distribution. Improvement in load balancing with WRCA assure better aggregation of sensor data and MAC layer performance. This is accomplished by incorporating a density center weighting parameter. This information is used to select density centered CHs. WRCA requires fewer messages to find density centered nodes as compared to TASC. Simulation results demonstrate the overall superiority in performance of the WRCA when mobility according to a realistic mobility model called Self similar Least Action Walk (SLAW) is considered. Simulation result shows that our proposed algorithm consumes 20% less energy and 50% more network lifetime as compared to WCA algorithm.