SVM-based analysis and prediction on network traffic

With continuous scale-up of the network and increase of the kinds of the services on the network, more and more people pay attention to the modeling and prediction for network traffic. Recently, SVM (Support Vector Machine), a new machine learning method, is comprehensively used to solve the problem of non-liner classification and regression. A network traffic predictive method presented in this paper is based on the LS-SVM (Least Squares SVM). Using NS2 simulator, we simulate the process of the network running with Drop-tail and RED controller respectively, then collect the being predicted traffic data which is on the bottleneck router. The results on the precision of prediction is good and feasible.