Network Traffic Prediction Based on Wavelet Transform and Genetic Algorithm

The traditional network traffic prediction is based on the establishment of a linear model, which can't describe the changes of network traffic accurately, resulting in low prediction accuracy. This paper proposes a new model of network traffic prediction based on wavelet transform and Genetic Algorithm. Firstly, after a wavelet decomposition, network traffic is turned into many stable components. Secondly, using BP neural network to predict each stable component, and optimizing neural network by genetic algorithm. Finally, all the prediction of components is combined to achieve highly-accurate traffic prediction. The experimental results show that the model has a better predictive effect.