• 中国计算机学会会刊
  • 中国科技核心期刊
  • 中文核心期刊

J4 ›› 2008, Vol. 30 ›› Issue (6): 29-32.

• 论文 • 上一篇    下一篇

校园网网络流量自相似性的测定

林兆启 林南晖 汪继东   

  • 出版日期:2008-06-01 发布日期:2010-05-19

  • Online:2008-06-01 Published:2010-05-19

摘要:

根据自相似性理论,网络流量自相似性的程度可由计算Hurst参数的大小来测定。本文通过在校园网主干路上进行网络数据包的采集,然后用方差时间方法和R/S方法计算所 采集流量的Hurst参数值,以检测校园网网络流量的自相似性。通过实验的分析与比较,验证了自相似性的存在,并对如何运用Hurst值的变化来进行流量异常检测做了初步的  探讨。

关键词: 网络流量 自相似性 方差时间方法 R/S方法

Abstract:

I he selt-simtlarity degree ot network trattic is checked by computing the Hurst parameter based on the selfsimilar theory. In this paper, we capture packets on the campus network's backbone, and then compute the Hurst parameter to check the self-similar degree of network traffic by using the R/S met    thod and the variance-time method. The existence of self-similarity is verified by experiments. We also discuss the network traffic abnormal detection b  ased on the Hurst parameter's obvious changes.

Key words: network traffic, self-similar, variance-time methods R/S method