CERNET流量行为季节预测模型

Seasonal Neural Network Model on Internet Traffic Behavior

网络流量行为预测是网络行为学的一个重要研究方向 .常规的网络流量预测大多采用的是 ARIMA时间序列模型 ,但普通时间序列预测模型的参数难以估计并且模型较难处理非平稳时间序列问题 .本文基于时间序列的神经网络模型研究 ,根据网络流量行为的季节性特点 ,提出了季节型神经网络模型 .用模型对 CERNET网络流量行为的预测分析表明 ,该模型预测效果较好 ,结果合理 ,对进行网络实时监控及网络管理都具有一定的理论和实践价值 .

The prediction on Internet behavior is an important face of network behaviorism. Traditional models on network traffic prediction are based seasonal ARIMA model, but it is difficult in finding its parameters and dealing with Non-stationary time series. Based on the neural-network model of time series and the season of network traffic behavior, a seasonal model of neural network is made by using artificial neural-network. At the time, the idea of making data smoothing process of Fourie before training is considered to improve the accuracy of prediction. The model was used in network traffic prediction of CERNET. The calculation results indicated that the model is reasonable and its accuracy is better than seasonal ARIMA model. It is valuable when being used in practices.