基于非参数CUSUM算法的DDoS攻击防御策略

On defending strategy for DDoS attack based on nonparametric CUSUM algorithm

分布式拒绝服务攻击(DDoS)具有突发性、攻击主机分布广等特点,对其防御尤显困难.从分析DDoS攻击原理及常见的防御策略入手,根据非参数CUSUM算法,给出了基于流连接密度时间序列分析的DDoS攻击防御策略及其模型,并进行了与理论值比较的实验取证.结果表明:该防御模型能够在很大程度上有效防御DDoS攻击.

Distributed Denial of Service （DDoS）has features including burstiness and wide distri_bution of attacked host which makes it difficult to defend it.Starting from analysis on the princi_ple of DDoS attack and common defending strategies,this paper gives defending strategy for DDoS attack based on time series analysis of flow connection density and the model according to nonparametric CUSUM algorithm,and carries out experimental evidence comparing with theo_retical values.Results show that the defending model can defend DDoS attack effectively to a large extent.