引用本文:孙知信,唐益慰,张伟,宫婧,王汝传.基于特征聚类的路由器异常流量过滤算法.软件学报,2006,17(2):295-304
【打印本页】   【下载PDF全文】   查看/发表评论  【EndNote】   【RefMan】   【BibTex】
←前一篇|后一篇→ 过刊浏览    高级检索
本文已被:浏览 5444次   下载 6883 本文二维码信息
码上扫一扫!
分享到: 微信 更多
基于特征聚类的路由器异常流量过滤算法
孙知信1, 唐益慰1, 张伟1, 宫婧1, 王汝传1
南京邮电大学,计算机科学与技术系,江苏,南京,210001
摘要:
基于当前入侵检测技术在检测到攻击的情况下没有良好的反应策略过滤攻击流量这一问题,提出了基于攻击流量特征聚类的特征提取算法AFCAA(anomaly traffic character aggregation algorithm).针对一般DOS(denial ofservice)/DDOS(distributed denial ofservice)攻击流数据包头中具有某些相似的特性,AFCAA通过运用重心原理进行统计聚类,在一定的欧氏距离范围内对基于目的IP的攻击流样本相应字段进行聚类划分,动态地提取出攻击流的重心作为攻击的特征.然后,及时地把其特征传输给Net Filter,可以进行高效的过滤,并保护正常流量的传输.实验结果表明,对当前流行的多种拒绝服务攻击,应用AFCAA系统的软件路由器都能够较准确地获取异常流量的特征,从而有效地进行过滤,减少攻击包传播的危害,保护有限的网络资源.
关键词:  拒绝服务(denial of service,简称DOS)  分布式拒绝服务(distributed denial 0f service,简称DDOS)  路由器  特征聚类  异常流量
DOI:
分类号:
基金项目:Supported by the National Natural Science Foundation of China under Grant No.60573141 (国家自然科学基金); the National High-Tech Research and Development Plan of China under Grant No.2005AA775050 (国家高技术研究发展计划(863)); the Scientific Research Foundation for the Returned Overseas Chinese Scholars, State Education Ministry (教育部回国人员基金资助), the Scientific Research Foundation for the Returned Overseas Chinese Scholars, Nanjing Government (南京市回国人员基金资助); the Scientific Research Foundation of Huawei Corporation (华为研究基金)
A Router Anomaly Traffic Filter Algorithm Based on Character Aggregation
SUN Zhi-Xin,TANG Yi-Wei,ZHANG Wei,GONG Jing,WANG Ru-Chuan
Abstract:
Under the situation of detecting attacks, current IDSs have no good reacting strategy to filter attack traffic. Based on network attacks’ traffic characters, an anomaly traffic character aggregation algorithm (AFCAA) is put forward. Because normal DOS (denial of service)/DDOS (distributed denial of service) attack traffic has some characters in their packets’ head, AFCAA uses the center of gravity theory to process statistic aggregation and aggregation partition based on the special field of the destination IP attack traffic in a fixed Euclid distance, and then it distills the center of attack traffic dynamically as the characters of attacks. Afterwards, through transmitting these characters to Net Filter, AFCAA can filter abnormal packets efficiently and protect the normal packet transmission. The experimental results show that the software router using AFCAA can efficiently find useful characters of prevalent DOS/DDOS attacks, reduce the harm of attack packets’ spreading, and protect the limited network resources.
Key words:  denial of service  distributed denial of service  router  character aggregation  anomaly traffic