Routing prediction for probabilistic mobility model using neural networks for ad-hoc networks

Abstract

Ad-hoc network is one of the key areas in wireless networks. There is always room for improvement in order to utilize the available resources optimally. One of the thought among these was to use the prediction algorithms to predict the next route of the mobile nodes which can help us in allocating the resources in advance and also controlling the network traffic. So this thesis mainly emphasis on the analysis of the behavior of AI based Routing Protocol for Adhoc networks suitable for video traffic Research has been done in this area using Neural Networks but no typical model was designed for predicting the routes due to lack of standard dataset to train the model. We generated different datasets using Gaussian Markov mobility model which has some concrete reasons or benchmarks to evaluate the end products' performance. Artificial Neural Networks (NN) & Extreme learning machine (ELM) were used to build the prediction models.