Botnet Forensic Analysis in Software Defined Networks using Ensemble Based Classifier

Software Defined Networks (SDN), a developing technology is replacing the conventional networking solutions that are currently in use. It helps greatly with network programmability and disassociates data plane from control plane. Due to its centralized administration, SDN is susceptible to number of threats. Botnet attacks are the most predominant kind of attacks. Peer-to-peer (P2P) botnets are a new, sophisticated variety of botnets. It is a decentralized network of computers that lack C & C servers and have been infected with malware and cooperate for the benefit of an attacker without the owners' permission. To combat botnet attacks, there are networking system solutions that use a single machine learning (ML) classifier, like the Support Vector Machine (SVM), decision tree (DT), etc. To create a more dependable and precise detection and mitigation application, detection speeds must be improved. In order to provide a system that effectively detects, classifies botnet traffic as Peer-to-Peer (P2P) and mitigates it, this paper uses extensive forensic analysis by combining SDN with decision tree and bagging integrated as ensemble machine learning classifiers to satisfy the necessary needs. This suggested method has 97 percent accuracy rate for identifying and mitigating botnet attacks on SDN networks.