适应频繁波动特点的移动话务量预测系统

Mobile Traffic Forecasting System Suits Frequent Fluctuations

移动通信话务量的时间序列具有较强的非线性和随机性,易受各种突发情况的影响。特别是在有频繁剧烈变动的情况下,很难用传统的预测方法进行预测。本文根据移动通信话务量的这一特点,采用分解及建模的方法,对话务量进行综合预测,经实践可以达到更好的效果。

Mobile communication traffic is nonlinear and radom. It＇s also often affected by all kinds of unexpected situations. Thus it＇s even more difficult to give an accurate prediction using conventional predicting methods when there＇s frequent dramatic changes. Based on the special character of mobile communication traffic, using division and modeling method on traffic analysing and predicting is proved to have a better effect.