Computer Engineering and Applications ›› 2014, Vol. 50 ›› Issue (4): 103-109.

Previous Articles     Next Articles

Network traffic prediction based on jointly optimization of embedding dimension and delay time

ZHANG Meng1,2, ZHANG Huyin1, YE Gang1   

  1. 1.School of Computer, Wuhan University, Wuhan 430072, China
    2.Network Center, Wuhan University, Wuhan 430072, China
  • Online:2014-02-15 Published:2014-02-14

延迟时间和嵌入维数联合优化的网络流量预测

张  萌1,2,张沪寅1,叶  刚1   

  1. 1.武汉大学 计算机学院,武汉 430072
    2.武汉大学 网络中心,武汉 430072

Abstract: In order to improve the prediction accuracy of network traffic, a network traffic prediction method is proposed based on jointly optimization embedding dimension(m) and delay time(τ) of phase space reconstruction according the relation between embedding dimension and delay time. Least squares support vector machine is used as the network traffic prediction algorithm and the optimal τ and m is selected according to prediction results of the network traffic, the simulation analysis is carried out on network traffic data to test the performance of single step and multi-step prediction model. The results show that the proposed method can effectively select the optimal τ and m, significantly improve the prediction accuracy of network traffic, the prediction results is significantly higher than reference methods of the network traffic.

Key words: network traffic prediction, phase space reconstruction, parameter optimization, least squares support vector machine, evaluation standard

摘要: 为了提高网络流量的预测精度,利用相空间重构的两个关键参数—延迟时间(τ)和嵌入维(m)间的相互联系,提出一种延迟时间和嵌入维数联合优化的网络流量预测模型。该模型以最小二乘支持向量机作为网络流量预测算法,根据网络流量预测结果优劣评价指选择最优[τ]和[m]值,建立单步、多步网络流量预测模型,并通过仿真实验对模型的性能进行分析。结果表明,模型可以准确选择出最优嵌入维数和延迟时间,显著提高了网络流量的预测精度,预测结果明显优于独立优化[τ]和[m]以及传统联合优化[τ]和[m]的网络流量预测模型。

关键词: 网络流量预测, 相空间重构, 参数优化, 最小二乘支持向量机, 评价标准