Exploiting Neural Networks for Mobility Prediction in Mobile Ad Hoc Networks

Abstract

The main aim of this paper is to predict the future movement of mobile nodes in adhoc networks using a neural network based method. This method consists of three-layer feedforwardnetwork. The backpropagation algorithm is used to learn the neural network. Frequent node mobilityin ad hoc networks causes unreliable wireless links. The mobility prediction allows to find the morestable links in a mobile ad hoc network. It reduces the overhead and improves the routing.The main aim of this paper is to predict the future movement of mobile nodes in adhoc networks using a neural network based method. This method consists of three-layer feedforwardnetwork. The backpropagation algorithm is used to learn the neural network. Frequent node mobilityin ad hoc networks causes unreliable wireless links. The mobility prediction allows to find the morestable links in a mobile ad hoc network. It reduces the overhead and improves the routing.