Localization of Wireless Sensors Networks Using Dynamic Cell Structures Neural Networks

Abstract

In Wireless Sensor Networks applications, the location phase is essential in order to accomplish the mission for which the network is deployed. In fact, several works have been carried out according to different research axes in order to improve more and more the accuracy of the localization, among others the use of the neural networks which have proved their contribution. This paper proposes and evaluates a Received Signal Strength-based localization using the Dynamic Cell Structures Neural Networks algorithm by comparing its performance with Multi Layers Perceptron Neural Network. All simulations are done in an indoor workspace and show that Dynamic Cell Structures is simple and efficient.