Human walk aware mobility resistant efficient clustering for data gathering in cell phone based wireless sensor networks

Recent research in human mobility patterns has shown truncated power law behavior for flight length and pause time distributions[8]. Various approaches have been applied to increase the efficiency of weighted clustering algorithms for mobile networks but no quantitative work has been done to exploit contextual mobility of human walk. In this paper we quantify the effect of human walk context through notions of super flight length and super pause time and uses them as parameters in the weighted clustering algorithm. We explore the premise that better stability of clustering can be achieved if the network is aware of super flight length and super pause time at node level. We demonstrate this for single-hop cellphone based sensor network where cellphone users generally exhibit truncated power law mobility characteristics. We are proposing three human walk context based Mobility Resistant Clustering Algorithm (HMRECA) which effectively captures human walk characteristics, and achieves better stability compared to WCA[7] of Mobile adhoc network and less power consumption compared to MRECA[6] algorithm of adhoc sensor networks. The context parameters used in HMRECA algorithms predicts the stability of clusters more effectively, compared to mobility parameter of WCA algorithm.