Visible to the public In-Design Resilient SDN Control Plane and Elastic Forwarding Against Aggressive DDoS Attacks

TitleIn-Design Resilient SDN Control Plane and Elastic Forwarding Against Aggressive DDoS Attacks
Publication TypeConference Paper
Year of Publication2018
AuthorsGillani, Fida, Al-Shaer, Ehab, Duan, Qi
Conference NameProceedings of the 5th ACM Workshop on Moving Target Defense
PublisherACM
ISBN Number978-1-4503-6003-6
KeywordsComputing Theory and Resilience, control theory, DDoS, pubcrawl, resilience, Resiliency, Scalability, SDN, SDN security
Abstract

Using Software-defined Networks in wide area (SDN-WAN) has been strongly emerging in the past years. Due to scalability and economical reasons, SDN-WAN mostly uses an in-band control mechanism, which implies that control and data sharing the same critical physical links. However, the in-band control and centralized control architecture can be exploited by attackers to launch distributed denial of service (DDoS) on SDN control plane by flooding the shared links and/or the Open flow agents. Therefore, constructing a resilient software designed network requires dynamic isolation and distribution of the control flow to minimize damage and significantly increase attack cost. Existing solutions fall short to address this challenge because they require expensive extra dedicated resources or changes in OpenFlow protocol. In this paper, we propose a moving target technique called REsilient COntrol Network architecture (ReCON) that uses the same SDN network resources to defend SDN control plane dynamically against the DDoS attacks. ReCON essentially, (1) minimizes the sharing of critical resources among data and control traffic, and (2) elastically increases the limited capacity of the software control agents on-demand by dynamically using the under-utilized resources from within the same SDN network. To implement a practical solution, we formalize ReCON as a constraints satisfaction problem using Satisfiability Modulo Theory (SMT) to guarantee a correct-by-construction control plan placement that can handle dynamic network conditions.

URLhttps://dl.acm.org/citation.cfm?doid=3268966.3268968
DOI10.1145/3268966.3268968
Citation Keygillani_-design_2018