Short-Term Forecasting: Simple Methods to Predict Network Traffic Behavior

摘要

In the article we evaluate the accuracy of simple linear forecasting methods applied to short-term prediction of network traffic behavior, namely the traffic intensity. Such investigation is carried out in order to determine the possibility of such methods employment in network management systems and various TE-implementations. Also time series extracted from real network traffic are statistically analysed to obtain general properties of aggregated network traffic behavior.