网络流量自相似性在DDoS攻击检测中的应用  

Application of Network Self-similarity in Detecting DDoS Attack

作　　者：杨永刚[1] 寇应展[1] 杨杰[1] 祝爱民[1] 

机构地区：[1]军械工程学院计算机工程系,石家庄050003

出　　处：《科学技术与工程》2007年第13期3296-3298,共3页Science Technology and Engineering

摘　　要：大量研究表明网络的正常流量具有自相似性,可以用自相似性参数Hurst系数来描述。DDoS攻击会破坏或者影响网络正常业务的自相似特性,用优化的R/S作为自相似性参数Hurst系数的估算算法,分析DDoS攻击数据流对正常业务数据流自相似参数Hurst系数的影响,来检测DDoS攻击。Amount of research has shown the network traffic with a feature of self-similarity. If there is DDoS attack, self-similarity of network traffic will change with it. Think of that, network traffic be can part changed with the normal one by analyzing of the parameter of network traffic. Then, the DDoS attack could be detected. Optimized R/S algorithm is used to rough estimate the parameter of self-similarity, namely Hurst value.

关 键 词：DDOS攻击 自相似性 自相似模型 网络自相似性 

分 类 号：TP393.02[自动化与计算机技术—计算机应用技术]