Defending DDoS attacks in software defined networking based on improved Shiryaev–Roberts detection algorithm

Article type: Research Article

Authors: Xiulei, Wang^a | Ming, Chen^a | Xianglin, Wei^{b; *} | Guomin, Zhang^a

Affiliations: [a] College of Command Information System, PLA University of Science and Technology, Nanjing, China. E-mails: [email protected], [email protected], [email protected] | [b] PLA University of Science and Technology, Nanjing, China. E-mail: [email protected]

Correspondence: [*] Corresponding author: Wei Xianglin, PLA University of Science and Technology, Nanjing, China. E-mail: [email protected].

Abstract: As the availability is an important issue of network security, it is necessary to research the DDoS defending mechanism of Software Defined Networking (SDN) before the real-world deployment. In this paper, we first measured the impact of DDoS attacks on SDN based on a test bed. The results showed that the resource utilization changed sharply when DDoS attacks started. Based on this observation, a lightweight DDoS defending mechanism which is implemented in SDN controller is put forward. In this mechanism, a score based Shiryaev–Roberts change point detection algorithm is firstly adopted to detect the starting time points of DDoS attacks. Then, the attacking flows will be filtered and rerouted to the middle boxes network where the DDoS attacking flows filtering algorithm is deployed. A prototype of the mechanism is built and evaluated in the test bed. The results show that the score based Shiryaev–Roberts detection mechanism can detect the attack quickly, and the controller along with switches can be protected from DDoS efficiently.

Keywords: Software defined networking, distributed denial of service, Shiryaev–Roberts detection algorithm, OpenFlow

DOI: 10.3233/JHS-150526

Journal: Journal of High Speed Networks, vol. 21, no. 4, pp. 285-298, 2015