基于节点随机采样的AS级IP追踪

An AS-level node probabilistic samping scheme for IP traceback

建立了PPM算法的通用数学模型,提出了一种新的基于节点采样的IP追踪方案。该方案采用新的标记信息编码机制,解决了传统包标记方案中由于地址分片带来的"组合爆炸"和误报率高的问题,标记过程以AS路径代替传统的IP路径,使用最优标记策略,使得路径重构过程具有更低的计算复杂性和更短的收敛时间。仿真分析表明,此方案具有应对大规模DDOS攻击源追踪的有效性和实时性。

A general mathematical model of PPM scheme is established.A novel probabilistic packet marking scheme for IP traceback is proposed,which is based on AS-level node probabilistic samping mechanism.In the scheme,using a new encoding scheme,we mitigate the high false positives incurred by combinatorial explosion significantly resulted by marking information fragment.More lower computational complexity and convergence time are obtained by reconstructing AS-level path instead of router-level path and applying optimal marking probability strategy.Correlation analysis suggest that our scheme is efficient and real-time to defend against large-scale DDoS attacks.