DDOS Attacks Detection and Mitigation in SDN using
Machine Learning
Journal:
GRENZE International Journal of Engineering and Technology
Authors:
Samuel Kennedy, R. Chitra, E. Vinodh Ewards, T. Jemima Jabaseeli, M. Sandhia
Volume:
10
Issue:
2
Grenze ID:
01.GIJET.10.2.659
Pages:
1972-1978
Abstract
A new networking paradigm called software-defined networking, or SDN, gives a
controller and its applications the capacity to see the whole network and to design it in a flexible
way. This allows for new developments in network protocols and applications. Many SDN
applications depend on the logically centralized control plane of SDN to offer the complete
network visibility, which is one of its main benefits. The literature offers a novel attack vector
peculiar to SDN that substantially undermine this basis. Our novel attacks have some spirit in
common with spoofing attacks in legacy networks (like the ARP poisoning attack), but they
vary greatly in that they take use of certain vulnerabilities that arise from the way
contemporary SDN functions differently from legacy networks. The knowledge about virtual
machines, which is a crucial building piece for both topology-aware SDN applications and core
SDN components, may be effectively poisoned by successful assaults.