An Efficient and Practical Defense Method Against DDoS Attack at the Source-End

Distributed Denial-of-Service (DDoS) attack is one of the most serious threats to the Internet. Detecting DDoS at the source-end has many advantages over defense at the victim-end and intermediate-network. One of the main problems for source-end methods is the performance degradation brought by these methods, which discourages Internet service providers (ISPs) to deploy the defense system. We propose an efficient detection approach, which only requires limited fixed-length memory and low computation overhead but provides satisfying detection results. The low cost of defense is expected to attract more ISPs to join the defense. The experiments results show our approach is efficient and feasible for defense at the source-end