Research on Combination Network Traffic Forecasting Model

With the continuous expansion of the scale of the Internet and the growing number of new network services, the network traffic data has complex behavioral characteristics. It brings a series of problems to effectively manage the network. Among them, how to establish an effective and accurate forecasting model is one of the hot research topics. For that reason, an improved GM (1,1) model is proposed to improve prediction accuracy. Then, to use the combination weighting method to combine the ARMA (p, q) model with the improved GM (1,1) model constitutes a new combination forecasting model. At last, campus network traffic prediction is based on the model. The results show the reliability and superiority of the combined forecasting model.