Improved R/S Algorithm Based on Network Traffic Self-Similarity

On the Internet, the distributed denial-of-service (DDoS) attack is the most frequently used and most deleterious attack which aims at a single target and causes denial of service for users of the targeted system. Today, network security has become the key for the popularization of all kinds of network applications. In this paper, a model against DDoS attack based on detecting the network traffic character is proposed and the method of detecting DDoS attack by computing the self-similarity parameter of network traffic is studied emphatically. The R/S algorithm is adopted to calculate the on-line Hurst parameter through data reusing, and then the implement is given. The experiment results show that this method can detect the DDoS attack accurately and in time.