摘要
采用了BP神经网络对网络流量数据的时间序列进行建模与预测。从分析网络流量的特征着手,构建了基于BP神经网络的IP网络流量预测模型,并进行了仿真验证。实验结果表明,该模型对网络流量的预测是有效可行的,并具有良好的收敛性和稳定性。
The time series of network traffic data was modeled and forecasted based on BP neural network . The IP network traffic forecasting model based on BP neural network was built by analyzing the characteristics of network traffic. The experimental results prove that the model is efficient in network traffic prediction with good astringency and stability.
出处
《南京邮电大学学报（自然科学版）》
2010年第2期16-21,共6页
Journal of Nanjing University of Posts and Telecommunications：Natural Science Edition
关键词
IP网络
流量预测
服务质量
BP网络
时间序列
IP Network
traffic prediction
quality of service
BP network
time series
