CSDSM: Cognitive Switch-based DDoS Sensing and Mitigation in SDN-driven CDNi Word
- 주제(키워드) SDN , CDN , CDNi , DDoS , Flash Crowd , Machine Learning , Support Vector Machine , Logistic Regression
- 주제(기타) Computer Science, Information Systems; Computer Science, Software Engineering
- 설명문(일반) [Mowla, Nishat I.; Doh, Inshil; Chae, Kijoon] Ewha Womans Univ, Dept Comp Sci & Engn, 51 Ewhayeodaegil, Seoul 03760, South Korea
- 관리정보기술 faculty
- 등재 SCIE, SCOPUS
- 발행기관 COMSIS CONSORTIUM
- 발행년도 2018
- URI http://www.dcollection.net/handler/ewha/000000151381
- 본문언어 영어
- Published As http://dx.doi.org/10.2298/CSIS170328044M
초록/요약
Content Delivery Networks (CDNs) are increasingly deployed for their efficient content delivery and are often integrated with Software Defined Networks (SDNs) to achieve centrality and programmability of the network. However, these networks are also an attractive target for network attackers whose main goal is to exhaust network resources. One attack approach is to over-flood the OpenFlow switch tables containing routing information. Due to the increasing number of different flooding attacks such as DDoS, it becomes difficult to distinguish these attacks from normal traffic when evaluated with traditional attack detection methods. This paper proposes an architectural method that classifies and defends all possible forms of DDoS attack and legitimate Flash Crowd traffic using a segregated dimension functioning cognitive process based in a controller module. Our results illustrate that the proposed model yields significantly enhanced performance with minimal false positives and false negatives when classified with optimal Support Vector Machine and Logistic Regression algorithms. The traffic classifications initiate deployment of security rules to the OpenFlow switches, preventing new forms of flooding attacks. To the best of our knowledge, this is the first work conducted on SDN-driven CDNi used to detect and defend against all possible DDoS attacks through traffic segregated dimension functioning coupled with cognitive classification.
more