A Coordination Of Extreme Learning Machine And ARMA Model Based On Wavelet Transform For Network Traffic Forecasting

Computer network has become one of the most important infrastructures of modern society in our daily life. At the same time, Internet penetration has increased dramatically, especially in some areas(such as China and India), this has become a great challenge for computer network security and management. Its chief purpose is to predict future network traffic by establishing a mathematical model for continuous network traffic data. If we could accurately obtain traffic data of the future, the network administrator would have analysed the possible problems and have taken appropriate measures in advance to guarantee maximum operation of the normal network. In addition, network traffic forecasting provides evidence of network anomalies or attack behaviours, and it is of great significance to the strengthening of network management, mastering traffic regular pattern, improving congestion service and load balancing.Its chief purpose is to predict future network traffic by establishing a mathematical model for continuous network traffic data. If we could accurately obtain traffic data of the future, the network administrator would have analysed the possible problems and have taken appropriate measures in advance to guarantee maximum operation of the normal network. In addition, network traffic forecasting provides evidence of network anomalies or attack behaviours, and it is of great significance to the strengthening of network management, mastering traffic regular pattern, improving congestion service and load balancing.