网络流量预测的建模与仿真研究

Simulation and Modeling of Network Traffic Prediction

研究网络流量准确预测问题,网络流量变化是一种具有时变性、多尺度和突发性的非线性系统,由于传统时间序列预测方法很难揭示内在变化规律,导致网络流量的预测精度比较低。为了提高网络流量的预测精度,提出一种小波分析BP神经网络的网络流量预测模型。模型首先通过小波分析对网络流量进行分解,得到网络流量信号的近似和细节部分,然后进行重构提取多尺度特征,最后将重构的网络流量数据输入到BP神经网络,利用BP神经网络的非线性能力对网络流量进行训练、建模并预测。仿真结果表明,小波神经网络方法提高了网络流量预测精度,是一种有效实用的网络流量预测方法。

Network flow changing is a system with multi-scale,complex and nonlinear,and it is difficult for traditional time series prediction method to reveal its internal change rules,so network traffic prediction accuracy is low.In order to improve the network traffic prediction precision,the paper proposed a network traffic prediction model based on wavelet analysis and BP neural network.Firstly,network traffic was decomposed by wavelet analysis to obtain the approximation and the detail signals.Then its multi-scale features were reconstructed and extracted.And lastly,the reconstructed network traffic data were input into the BP neural network.The experiment results show that the wavelet neural network algorithm can improve the network traffic prediction accuracy,and it is a practical and effective network traffic prediction method.