Back propagation neural network based lifetime analysis of wireless sensor network

Abstract

As large amount of energy constrained nodes are randomly distributed in network, the entire lifetime of wireless sensor network is difficult to estimate. In this paper, a back propagation (BP) neural network based Markov model is presented to calculate the lifetime of wireless sensor network. BP neural network is employed to reduce the calculation difficulty of Markov state equation. The simulation results indicate that this method gives the value of maximum lifetime exactly and its computing complexity is low. The quantitative degree of computing reliability relative error between the results of three layer BP neural network and the Markov model is 10-4. © 2009 Springer Berlin Heidelberg.