A noble approach for self learning and cluster based routing protocol with power efficiency in WSN

Energy efficiency is the central issue for developing a routing protocol for wireless sensor network. We propose a statistical model for a self learning, stable clustering power efficient routing protocol. The algorithm uses statistical functions like mean, variance and standard deviation to imprecise the data to be sent to the base station and the threshold value for generating an alarm during emergency. The proposed clustering protocol exploits the statistical similarity of the sensed environmental data to explore any emergency or unusual value in the currently sensed data and automatically alerts the base-station about it. The algorithm also solves the contemporary problem of corresponding generation of both periodic and event driven data. Finally a simulation is done to validate the results pertaining to the improvement in power efficiency.