Real-Time Network Traffic Prediction Based on a Multiscale Decomposition

The presence of the complex scaling behavior in network traffic makes accurate forecasting of the traffic a challenging task. Some conventional prediction tools such as recursive least square method do not apply to network traffic prediction. In this paper we propose a multiscale decomposition approach to real time traffic prediction. The raw traffic data is first decomposed into multiple timescales using the à trous Haar wavelet transform. The prediction of the wavelet coefficients and scaling coefficients are performed independently at each timescale using ARIMA model. The predicted wavelet coefficients and scaling coefficient are then combined to give the predicted traffic value. This multiscale decomposition approach can better capture the correlation structure of traffic caused by different network mechanisms, which may not be obvious when examining the raw data directly. The proposed prediction algorithm is applied to real network traffic. It is shown that the proposed algorithm generally outperforms traffic prediction using neural network approach and gives more accurate result. The complexity of the prediction algorithm is also significantly lower than that using neural network.