Prediction of Distributed Denial of Service Attacks in SDN using Machine Learning Techniques

A network architecture known as "software-defined networking" (SDN) enables the design and creation of hardware elements remotely. Due to their permanent connections, you can modify on the go. Even if SDN remains fantastic approach, Internet-threatening DDoS occurrences may exploit it. DDoS occurrences may be stopped with the use of machine learning techniques. When numerous systems work together to assault a server, it is called a DDoS attack. A control layer software manages the devices of the infrastructure layer and enables SDN by linking the application layer to the infrastructure layer. [1] The methods for identifying malicious traffic that we discuss in this work include decision trees, Cat Boost, and Extra trees. The outcomes of our tests demonstrate that Decision Tree, Cat Boost, and Extra Tree are capable of determining if the attack is safe or not.