A Novel Technique to Control the Traffic of Wireless Ad-hoc Network by Fuzzy Systems and Prediction with Neural Network

Abstract

Wireless Ad-hoc Networks have been proposed for variety of applications. In this paper our goal is to find crisis points in the roads, such as accident locations, and asphalt-destruction points, which may cause traffics along the roads. The applied method for traffic prediction is based on back propagation algorithm. In this paper the length of packets are assumed to be constant. Along the road the traffic is modeled by a Poisson Process. With this model we can produce packets. The number of packets which are sending from a typical node to the other nodes is controlled by a fuzzy system. By appropriate training of a neural network, we have shown that, having the traffic packets at time (t) for each node, we can predict the traffic of each node at next time successfully.