A SYN Flood Detection Method Based on Self – similarity in Network Traffic

Since the normal data fail to be transmitted under the SYN Flood attack, the paper proposes a detection method which can rapidly and accurately detect the SYN Flood attack. First, it takes a real - time intercept of network traffic, and selects network traffic to discrete. Second, the fitting function can be achieved by fitting the discrete network traffic repeatedly. Finally, the integral value of the fitting function is calculated, which is used to compare with the Hurst value. The SYN Flood attack can be effectively detected by comparing the integral value, which calculated by the fitting function curve, with the Hurst value of the network traffic.