基于季节变动模型的网络流量行为预测

Forecasting Network Traffic Behavior Based on Seasonal Fluctuant Model

网络的流量行为特征在一定程度上反映了网络运行的某种规律,对流量行为的分析和预测不仅有助于对网络的整体理解,还可为网络管理提供决策支持。本文研究了确定时间序列方法中的季节变动模型,并将其应用于校园网的日流量预测。通过比较实际流量的抽样值和预测值,证明了预测模型具有较高的可信度。

The characters of network traffic behavior reflect some rules of network running. The analysis and forecast of traffic behavior can not only make it easier to understand network, but also support decision-maklng in managing network. The seasonal fluctuant model of assured time series is studied and applied to forecast daily traffic in campus LKN. After comparing actual traffic samples with forecast result, the forecast model is proved reliable.