A Short-Term Forecasting Algorithm for Network Traffic Based on Chaos Theory and SVM
Journal of Network and Systems Management
Hi-index | 0.00 |
IP traffic prediction plays an important role in network-layout, traffic-management, as well as the emphasis of traffic-project, congestion-control and network management. Poor prediction performance would be acquired generally as a result of intense nonlinearity of networks traffic. To tackle it, a modeling method for exact representing IP traffic’s movement tendency and a regression algorithm with powerful nonlinear approaching ability should be employed. Consequently, Chaos theory and Support Vector Machine (SVM) win the bid. Then, an improved algorithm based-on Local SVM method for small scale data-set is proposed. Experimental results demonstrate the validity of improvement by a real-life paradigm that successful forecasting with continuously daily IP traffic during a few days gathered from campus network.