Network traffic prediction based on ARFIMA model

Date Added: Nov 2012
Format: PDF

ARFIMA is a time series forecasting model, which is an improve d ARMA model, the ARFIMA model proposed in this paper is demonstrated and deduced in detail. combined with network traffic of CERNET backbone and the ARFIMA model, the result shows that, compare to the ARMA model, the prediction efficiency and accuracy has increased significantly, and not susceptible to sampling. Network traffic prediction is an important application of network management and network measurement research direction, network traffic prediction model is the model to predict trends of sometime things in the future under the theory guidance.