SD-Anti-DDoS: Fast and efficient DDoS defense in software-defined networks

Yunhe Cui, Lianshan Yan, Saifei Li, Huanlai Xing, Wei Pan, Jian Zhu, Xiaoyang Zheng

Abstract: In order to overcome Distributed Denial of Service (DDoS) in Software Defined Networking (SDN), this paper proposes a mechanism consisting of four modules, namely attack detection trigger, attack detection, attack traceback and attack mitigation. The trigger of attack detection mechanism is introduced for the first time to respond more quickly against DDoS attack and reduce the workload of controllers and switches. In the meantime, the DDoS attack detection method based on neural network is implemented to detect attack. Furthermore, an attack traceback method taking advantages of the characteristics of SDN is also proposed. Meanwhile, a DDoS mitigation mechanism including attack blocking and flow table cleaning is presented. The proposed mechanism is evaluated on SDN testbed. Experimental results show that the proposed mechanism can quickly initiate the attack detection with less than one second and accurately trace the attack source. More importantly, it can block the attack in source and release the occupied resources of switches.