DDoS Detection and Mitigation SDN using Support Vector Machine
S Prajwal, M Siddhartha, Sai Charan, L, Girish · Zenodo (CERN European Organization for Nuclear Research) · 2021
In recent years, the rise of software-defined networks (SDN) has made network control more flexible, easier to set up and manage, and has provided a stronger ability to adapt to the changing demands of application development and network conditions. The network becomes easier to maintain but also achieves improved security as a result of SDN. The architecture of SDN is designed for Control Plane and Forwarding Plane separation and uses open APIs to realize programmable control. SDN allows for the importing of third-party applications to improve network service or even provide a new network service. In this paper, we present a defense mechanism, which can find attack packets previously identified through the Sniffer function, and once the abnormal flow is found, the protection mechanism of the Firewall function will be activated. For the capture of the packets, available libraries will be used to determine the properties and contents of the malicious packet and to anticipate any possible attacks. Through the prediction of all latent malicious behaviors, our new defense algorithm can prevent potential losses like system failures or crashes and reduce the risk of being attacked.