Network Traffic Chaotic Forecasting Based on Jointly Solving Forecasting Parameters of Model

Network traffic forecasting is the basis of network management, in order to improve the network traffic forecasting precision, considering the mutual influence between parameters of model, so this paper put forward a network traffic chaos forecasting model based on jointly solving forecasting parameters of model. Firstly, network traffic history data are collected and use chaotic theory reconstruct historical data, and determine the scope of the model parameters, and secondly, genetic algorithm is used to determine most reasonable parameters according to optimal individuals based on nature "the survival of the fittest, superior bad discard" mechanism, finally, the most reasonable parameters are used to establishing network traffic forecasting model by using training sample of network traffic, and the simulation experiments is used to analyze performance of the mode. Results show that the proposed model can accurately mine network traffic network flow change characteristics among historical data by chaos theory and get high forecasting accuracy of network traffic forecasting, it provides a modeling tool for nonlinear network traffic forecasting.