A Machine Learning Approach for Predicting DDoS Traffic in Software Defined Networks

Abstract

Software Defined Networks (SDN) paradigm was introduced to overcome the limitations of the traditional network such as vendor dependencies, inconsistency policies, etc. It becomes a promising network architecture that provides the operators more control over the network infrastructure. The controller also called the operating system of the SDN has the centralized control over the network. Despite all its capabilities, the introduction of various architectural entities poses many security threats to SDN layers. Among many such security issues, Distributed Denial of Services (DDoS) is a rapidly growing attack that poses a tremendous threat to SDN. It targets to the availability of the network, by flooding the controller with spoofed packets. It causes the controller to become paralyzed, and thereby the entire network becomes destabilize. Therefore, it is essential to design a robust DDoS detection mechanism to prevent the control plane attack. In this regard, we have used seven Machine Learning techniques to accurately classify and predict different DDoS attacks like Smurf, UDP flood, and HTTP flood. Experimental results with proper analysis have been presented in this work.