Automatic model classification of measured Internet traffic

A new method for real-time traffic model classification is proposed and evaluated. The method classifies the current measured traffic to a "best-fit" model selected from a library of candidate models using statistical estimation techniques. A simple two-model system has been prototyped and evaluated through simulation experiments. The experimental system consists of a short-range dependent model and long-range dependent model, and uses the estimated Hurst parameter to select between the two models. Results demonstrate that the two-model system can classify observed traffic to the correct model with fair accuracy, and can automatically detect a change in traffic characteristics after a delay. The design parameters affecting the classification accuracy and the delay to detect traffic changes are discussed.