Intelligent Traffic Matrix Estimation Using LevenBerg-Marquardt Artificial Neural Network of Large Scale IP Network

Abstract

The expansion in computer networks leads the traffic matrix estimation problem to be an essential component in managing the networks. In this paper, we have estimated the large IP network traffic matrix (TM) using a Levenberg-Marquardt Neural Network. Traffic matrix estimation is generally a complicated task for a large scale IP network, therefore, we involved neural networks for the intelligent and accurate estimation of it. The performance of proposed Neural Network has been verified using Abilene dataset available publicly for researchers and it has been observed that our used algorithm gives promising results when it comes to traffic matrix estimation.