摘要
针对当前网络流量预测法是通过监测网络流量历史数据进行预测,存在预测精准度低和流量信息参数自适性差的问题,提出基于多元线性回归分析的高峰期网络流量预测模型。通过BP神经网络法,确定网络流量信息权值,采用滑动窗口算法得到流量序列中对应信息数据,构成新的网络流量序列,得到多元线性回归初始模型;引入最小二乘法对流量信息参数进行估算,得到流量信息的样本回归函数,使用可决系数F检验及统计样本回归函数,完成高峰期网络流量预测模型的构建。实验结果表明,使用该模型可降低误差、提高拟合度、增加能量利用率,为高峰期网络流量预测提供了基础保障。
In view of the current network traffic prediction method is predicted by the historical data flow monitoring network, prediction accuracy is low and the flow of information parameter adaptive problem of poor peak network traffic prediction model is proposed based on multiple linear regression analysis. The BP neural network method to determine the weights of network traffic information, using the sliding window algorithm to obtain the corresponding information data flow sequence, a new network traffic sequence, the multiple line- ar regression model to obtain the initial; by the least squares method to estimate the parameters of traffic information, get traffic infor- mation of the sample regression function, using the coefficient of determination F test and statistics the regression function, complete the peak network traffic prediction model construction. The experimental results show that the model can reduce the error, improve the fitting degree and increase the energy utilization rate, and provide the basic guarantee for the peak network flow prediction.
作者
刘维嘉
LIU Weijia1,2(Yangling Vocational and Technical CollegeYangling, Yangling, 712100, Chin)
出处
《网络新媒体技术》
2018年第2期41-47,共7页
Network New Media Technology
关键词
高峰期
网络流量
预测模型
多元线性回归
流量序列
回归函数
Peak period
network traffic
prediction model
muhiple linear regression
flow series
regression function
