Wavelet Neural Network Based Moving Object Tracking for Wireless Sensor Network

Abstract

This paper presents an intelligent moving object tracking strategy for Wireless sensor network with more real time scenario. Neural network based estimation framework has been proposed in this work to track the node in Wireless sensor network. Nodes localization has emerged as a major challenge in many WSN applications where the instantaneous location of sensor node governs the power and cost at physical layer and network layer. The phenomenon like reflection, multipath fading, uncertain environmental dynamics etc makes the problem more severe. The localization accuracy is further reduced when the nonlinear sensor characteristics play the role during received signal strength measurement. Neural network based intelligent node localization scheme is proposed in this work where wavelet function has been used as activation function to enhance the robustness and accuracy of the technique. Online tuning laws are derived in this paper for wavelet neural network training and the accurate target tracking. The performance of the proposed technique is evaluated using a simulation study.