¬Network Traffic Prediction Using the Combination of Chaos Theory and Simple-MKL

Network traffic prediction has been an important problem in the field of network designing and planning. Several prediction approaches are developed to predict the network traffic as well as its generalization promotion ability. By analyzing the limitation of previous approaches, we propose a network traffic prediction approach based on the Chaos theory and Regression-SMKL (RRSMKL) algorithm. To our best knowledge, this is the first work that couples Chaos theory and RSMKL for network traffic prediction. First, we employ Chaos theory to extract the features of networks. Then we propose the RSMKL is to predict the network traffic, in the manner of regression. The proposed approach is then applied to predict the real network traffic data. By the simulation test, the results show that the proposed prediction approach in this paper has satisfied prediction accuracy, fast prediction speed and better network traffic prediction ability.