基于小波方法的Internet流量的预测建模  

Forecasting Model of Internet Flow Based on Wavelet Transform

作　　者：曹雪[1] 魏恒义[1] 程竹林[1] 曾明[1] 

机构地区：[1]西安交通大学计算机科学与技术系,西安710049

出　　处：《计算机工程》2003年第15期56-57,114,共3页Computer Engineering

基　　金：国家高技术研究发展计划资助项目(2001AA112111)

摘　　要：小波模型是自相似过程的流量模型， Internet流量数据属于非平稳的时间序列，小波变换可将非平稳的时间序列变为多个平稳的分量，再对分量采用相应的回归模型进行预测，然后将各个预测分量利用小波重构成最终的预测流量。通过实例具体说明了如何利用小波变换对Internet流量数据进行分析、预测。Wavelet model is a self-similar model. The Internet traffic belongs to non-stationary time series. Wavelet transform can decompose non-stationary time series into several stationary components, and then all these components are forecasted by relevant regression model. Subsequently, the forecasted traffic is formed by wavelet reconstruction with the forecasted components. Finally, how to analyze and forecast network traffic through utilizing wavelet transform is demonstrated through an example.

关 键 词：小波变换 非平稳时间序列 流量预测 

分 类 号：TP393.06[自动化与计算机技术—计算机应用技术]