Fault Detection of Large Scale Wireless Sensor Networks using Six Sigma score

Abstract

Technological developments in wireless sensor networks (WSN) through the use of various networking protocols, it provides enough scope in vast range of applications of the physical world that guarantees for consistency and accuracy of data. With the help of six sigma score and correlation as a parameter, faulty sensor nodes are identified. Using Markov Chain in the network, rank/rating was found from the correlation. From the light of Rating, a statistically efficient algorithm was developed known as Six Sigma z-score. Simulation results show that it is effective in finding faulty nodes giving importance to quality nodes. Experimental results show it outperforms both voting and statistical algorithms test in terms of detection accuracy and false alarm rate.