Application of Fuzzy AR in Time-series Decomposed Model of Network Traffic
Han Weihuan · Jisuanji gongcheng · 2004
Traditional time-series decomposed model applies AR model to the residual,which is obtained by decomposing trend component and period component from the network traffic. Accordingly,the model can only handle the large scale traffic,but is unable to model the long-dependence traffic in small scale. By replacing AR model with fuzzy AR model,the improved time-series decomposed model can handle the network traffic in all kinds of time scale.