Proposal of a new approach for efficient intrusion detection systems in SDN networks

Abstract

In the fast-changing field of network security, Software-Defined Networking (SDN) stands out as a game-changer by providing centralized control and programmability of network resources, which leads to more flexible and efficient management. However, this centralization also brings new vulnerabilities, making SDN environments appealing targets for cyber-attacks. To address these risks, Intrusion Detection Systems (IDS) are essential, as they monitor and analyze network traffic to detect and respond to malicious activities. Traditional Intrusion Detection Systems (IDS) that depend on centralized data collection and processing encounter major privacy and scalability issues. As network environments become more complex and data volumes increase, these systems often fail to meet modern requirements, resulting in potential single points of failure and heightened privacy risks. This project tackles these challenges by investigating the use of machine learning (ML) and deep learning (DL) models to enhance IDS within Software-Defined Networking (SDN) environments. ML and DL methods can greatly boost the accuracy and efficiency of intrusion detection by analyzing extensive datasets and detecting patterns that suggest malicious activities. This project focuses on a comprehensive dataset analysis, applying diverse machine learning (ML) and deep learning (DL) models, and evaluating their performance in detecting intrusions within Software-Defined Networking (SDN) environments. The main objective is to improve the effectiveness of Intrusion Detection Systems (IDS) in SDN by utilizing advanced models, recognizing their strengths and weaknesses, and contributing to the creation of more robust, scalable, and privacy-preserving IDS solutions. The insights obtained will ultimately enhance the security of SDN environments.