Network Traffic Forecasting Based on Wavelet Analysis and Relevance Vector Machine

Aiming at chaotic characteristics of network traffic and the shortage of traditional forecasting models, In order to improve forecasting accuracy of network traffic, a new network traffic prediction model based on chaos theory, wavelet analysis and relevance vector machine is proposed in this paper. Firstly, wavelet analysis is used to decompose the network traffic and the components of different frequency characteristics are obtained, and secondly, phase space reconstruction of components is carried out by using chaos theory, lastly, wavelet analysis is used to combine and get the final results of network traffic and network traffic is used to do simulation experiment. The results show that the compared with other models, the proposed model can accurately reflect the chaotic characteristics of the network traffic, and obtain the higher accuracy of the prediction results.