Detection of DDoS Attacks Against Wireless SDN Controllers Based on the Fuzzy Synthetic Evaluation Decision-making Model
Qiao Yan, Qingxiang Gong and Fang-An Deng
Software Defined Networking (SDN) is a new network architecture that separates the control plane and the data plane and provides logically central control over the whole network. Because SDN controller combines the upper application layer and the underlying infrastructure layer, it may face the problem of single-point failure. If it is made unreachable by a Distributed Denial of Service (DDoS) attacks, the whole network may not work normally. Especially for wireless SDN controllers, due to the secure channel for the control protocol in communication between wireless SDN controller and wireless SDN devices is exposed in the attacker’s field of vision, the attack range of DDoS attackers will be expanded. To mitigate this threat, this paper introduces a solution based on fuzzy synthetic evaluation decision-making model that is effective and lightweight in terms of the resources that it uses. Importantly, it takes many factors that can be used to detect DDoS attacks into consideration and makes a comprehensive judgment according to multifactors. To test the solution, the paper also proposes three kinds of DDoS attacks specialized for SDN network and presents two kinds of DDoS attacks inherited from traditional network. Every attack has been tested with the detection method. Finally, we also make a comparable experiment to show its advantage to other DDoS detection algorithm based on single factor. The results show its efficiency in detecting most of the DDoS attacks.
Keywords: SDN, openflow, DDoS attacks, fuzzy synthetic evaluation decision making model, entropy