Modelling and Prediction of Internet Time-Delay by Feed-Forward Multi-layer Perceptron Neural Network

Abstract

Internet time-delay prediction plays a key role in improving dynamic performance for many applications, especially for Internet-based tele-operation systems. Therefore, ideal prediction approach for Internet time-delay must be investigated emphatically. In this paper, after analayzing Internet time-delay, a new approach based on neural network is developed to predict the uncertain time-delay in the Internet. By using measured data between four Internet nodes and MATLAB environment, the feed-forward multi-layer perceptron (FMLP) neural network has been trained with single-step-ahead Prediction algorithm. Then, using validation data, the performance of the neural network was evaluated. It is shown that this neural network can predict time-delay based on their proper inputs.