设计并实现了1种基于小波变换的DDoS(分布式拒绝服务)攻击快速在线检测器.采用自定义的Socket方式进行在线数据包捕获,采用链表形式的数据包特征库进行数据特征存储,利用小波变换求解自相似系数在线检测DDoS攻击.试验结果表明,该设计可以快速区分DDoS攻击与正常大流量数据,并可有效检测出非知名DDoS攻击工具发起的攻击.
A fast online detection method aiming at DDoS(distributed denial of service) attack detection based wavelet transform was presented. The user-defined Socket method was adopted to capture network packets, the linked-list packets character library was responsible for keeping packets, and DDoS attack was detected online based on changes in the characteristic of network self-similarity which was estimated by using the real-time wavelet transform. The experiment results show that the detection can distinguish real DDoS attack from the massive data flow, and can effectively detect DDoS attack launched by unknown DDoS tools fast.