Time Series Applied to Network Traffic Prediction: A Revisited Approach

S. Fernandes, V. Teichrieb, D. Sadok, and J. Kelner (Brazil


communications and networks, network traffic prediction, time series modeling


This paper presents a revisited approach to apply time series modeling and analysis to network traffic modeling and prediction. This approach extends the use of classical procedures in predictability analysis of network traffic by using a combination of classical techniques such as the Exponential Smoothing and the Box & Jenkins methods. The obtained results are based on real traces captured from intradomain backbone network traffic.

Important Links:

Go Back