Network Traffic Prediction Based on Neural Network

Predicting and modeling network traffic is always an important subject in network capability studying. The aim of this article is to explore a new network model in order to describe and predict the network character accurately. Firstly wavelet neural network is investigated and its disadvantages are analyzed. In order to overcome disadvantages of wavelet neural network, genetic algorithm is used to optimize weight and threshold of neural network. At last, the proposed algorithm is used in network traffic prediction and the results show the proposed scheme has good performance.