Comparison of Network Intrusion Detection System for Software-Defined Networks Using Tree-Based Classifiers
Pogiri Deepika, Ganesh S. Pritham, S. Barath Krishna, R. Rajasimman · 2024
With the world moving towards automation, networks and their functions are no different. The advent of Network Function Virtualization (NFV) to provide basic networking functions and Software-Defined Networks (SDN) to orchestrate overall operations of the network has enabled more automated provisioning and policy-based management of network resources. Though these technologies increase efficiency and support automation, they still fall prey to online adversaries and attacks such as DDoS, XSS, brute force attacks, SQL injection attacks, etc. In this article, we provide a novel way to detect network intrusions by malicious threat actors that attempt to gain access or compromise the system using tree-based classifiers. The system is capable of detecting multiple types of attacks and also attacks that are similar. Using machine learning algorithms like random forest algorithm and XGBoost, which are based on trees, to make use of multiple features to detect malicious activity inside the network. Keywords : Artificial intelligence, Intrusion detection, Random forest, XGBoost