基于Bayes分类器的DoS/DDoS攻击防御技术  

A Defense Technique against DoS/DDoS Attack Based on Bayes Classifier

作　　者：罗建[1] 李艳梅[1,2] 

机构地区：[1]西华师范大学计算机学院,四川南充637002 [2]电子科技大学计算机科学与工程学院,四川成都610054

出　　处：《航空计算技术》2010年第2期127-129,134,共4页Aeronautical Computing Technique

基　　金：国家863计划项目资助(2008AA01Z132)

摘　　要：提出了一种基于贝叶斯分类算法的分布式拒绝服务攻击防御技术。利用贝叶斯分类算法来计算数据包的分值特征,按照分值表对数据包进行评分并映射成危险等级。然后对危险等级进行评估,根据不同危险等级对网络流量进行过滤。基于该防御技术,设计并实现了防御系统。通过实验分析,该系统在DoS/DDoS攻击发生时能有效区分正常流量与异常流量,从而实现对DoS/DDoS攻击进行实时防御。A defense technique against DoS/DDoS attack based on Bayesian classification algorithm is proposed. The technique uses Bayes classifier to calculate the value characteristics of the data packets, it evaluate the risk score and map the danger level in accordance with the score table. And then cleanses the network traffic according as the danger degree of the packets. We design and implement defense system based on the technique. It is proved by the experiment that this system can effectively differentiate between normal traffic and abnormal traffic,which can take real- time defense against DoS/DDoS attack.

关 键 词：拒绝服务 分布式拒绝服务 贝叶斯分类器 危险等级 包过滤 

分 类 号：TP393.08[自动化与计算机技术—计算机应用技术]