基于随机包标记的不重复标记追踪模型*

DDoS攻击传统的防御措施包括如防火墙、入侵检测系统等采取的是被动防御的策略，防御效果不够理想。采用主动防御策略的攻击源追踪技术，提出一种新的攻击源追踪模型，不需要AMS算法所要求的受害者预先具备上游拓扑数据的强假设前提。新的标记算法对标记过边信息的包不再重复标记，通过上游路由器的配合确认就能定位攻击源，与AMS算法和改进后的AEMS算法相比收敛速度更快，受标记概率和路径长度的影响更小、更稳定。

Non-repeatedly mark model based probabilistic packet marking for IP traceback

The existing traditional countermeasures in defending against distributed denial-of-service (DDoS) attacks,such as firewall and intrusion detection system(IDS),can not do well only by passive defense policy.The essay pointed out a new IP traceback model based on probabilistic packet marking, this model inherited the advantages of AMS,but it did not need the assumptions of having upstream topology by victim.The new algorithm did not mark the packet of having edge message again,it was much faster in convergence time and more stable, and had less influence by marking probability and the length by routing.