Control and Prediction of Communications Network Congestion by Using Artificial Intelligent techniques

There exists great demand supporting ever-increasing new Internet applications such as voice over IP and video on demand.This causes network congestion, which occurs when the load on the network is greater than the capacity of the network, despite of the implementation of various conventional control algorithms, congestion remains a critical concern in network systems. Hence it is necessary to find some non-conventional intelligent techniques to control and predict congestion. In this work, intelligent Control Techniques (ICT), specifically Fuzzy Logic Control (FLC) is utilized to control congestion and keep the load lower than the network capacity. We propose using learning techniques (WEKA )the concept from another intelligent control technique which is the Artificial Neural Networks to predict network congestion problems before they start impacting the performance of services. We trained the network so it would not stick with the conventional "either or" logic, but also consider the cases close to both, then we tested the network using data known to us but not to the network. Finally, we evaluated the system by hitting the network with completely strange to it and to us data and see if it would accomplish the job it was trained for. also we used OPNET Modeler, for network simulation and final results validation.