Multiple-Step-Ahead Traffic Prediction in High-Speed Networks

Traffic in high-speed networks shows distinct patterns at different timescales. This characteristic should be taken into account to address the error propagation in the multiple-step-ahead traffic prediction. Based on this idea, we proposed an algorithm in which traffic is modeled at different timescales using Gaussian process regression (GPR). The prediction at a timescale is made using the data of that timescale as well as the prediction results at larger timescales. Experiments performed on two public traffic data sets show that our algorithm has lower error propagation than other algorithms, including ARIMA, FARIMA, LSTM, and Convolutional LSTM.

[1]  Amir F. Atiya,et al.  A review and comparison of strategies for multi-step ahead time series forecasting based on the NN5 forecasting competition , 2011, Expert Syst. Appl..

[2]  Michalis Faloutsos,et al.  Long-range dependence ten years of Internet traffic modeling , 2004, IEEE Internet Computing.

[3]  Rainer Palm,et al.  Multiple-step-ahead prediction in control systems with Gaussian process models and TS-fuzzy models , 2007, Eng. Appl. Artif. Intell..

[4]  Hong Liu,et al.  Predicting Inter-Data-Center Network Traffic Using Elephant Flow and Sublink Information , 2016, IEEE Transactions on Network and Service Management.

[5]  Walter Willinger,et al.  Long-Range Dependence and Data Network Traffic , 2001 .

[6]  C. Rasmussen,et al.  Gaussian Process Priors with Uncertain Inputs - Application to Multiple-Step Ahead Time Series Forecasting , 2002, NIPS.

[7]  Mohamed Cheriet,et al.  Gaussian Process Regression Based Traffic Modeling and Prediction in High-Speed Networks , 2016, 2016 IEEE Global Communications Conference (GLOBECOM).

[8]  Jiankun Hu,et al.  Modeling Oscillation Behavior of Network Traffic by Nested Hidden Markov Model with Variable State-Duration , 2013, IEEE Transactions on Parallel and Distributed Systems.

[9]  Z. Sun,et al.  Traffic predictability based on ARIMA/GARCH model , 2006, 2006 2nd Conference on Next Generation Internet Design and Engineering, 2006. NGI '06..

[10]  Dit-Yan Yeung,et al.  Convolutional LSTM Network: A Machine Learning Approach for Precipitation Nowcasting , 2015, NIPS.