Review Paper on Predicting Network Attack Patterns in SDN using ML

Software Defined Networking SDN provides several advantages like manageability, scaling, and improved performance. SDN has some security problems, especially if its controller is defense less over Distributed Denial of Service attacks. The mechanism and communication extent of the SDN controller is overloaded when DDoS attacks are performed against the SDN controller. So, as results of the useless flow built by the controller for the attack packets, the extent of the switch flow table becomes full, leading the network performance to decline to a critical threshold. The challenge lies in defining the set of rules on the SDN controller to dam malicious network connections. Historical network attack data are often wont to automatically identify and block the malicious connections. In this review paper, we are going to propose using ML algorithms, tested on collected network attack data, to get the potential malicious connections and potential attack destinations. We use four machine learning algorithms C4.5, Bayesian Network BayesNet , multidimensional language DT , and Naive Bayes to predict the host which will be attacked to support the historical data. DDoS attacks in Software Defined Network were detected by using ML based models. Some key features were obtained from SDN for the dataset in normal conditions and under DDoS attack traffic. Dr. C. Umarani | Gopalshree Kushwaha "Review Paper on Predicting Network Attack Patterns in SDN using ML" Published in International Journal of Trend in Scientific Research and Development (ijtsrd), ISSN: 2456-6470, Volume-4 | Issue-6 , October 2020, URL: https://www.ijtsrd.com/papers/ijtsrd35732.pdf Paper Url: https://www.ijtsrd.com/computer-science/computer-network/35732/review-paper-on-predicting-network-attack-patterns-in-sdn-using-ml/dr-c-umarani