Doctorate thesis defense of Amina SAHBI

Doctorate thesis defense on March 26^th 2026 at 09H00 AM ,in Amphitheater Ibn Khaldoun, SUP'COM 2.

Entitled : AI-driven Security Incident Detection in SDN/NFV Networks

Presented by : Amina SAHBI

Committee

Abstract

Modern communication networks increasingly rely on Software-Defined Networking (SDN) and Network Function Virtualization (NFV), which not only enhance flexibility and scalability but also introduce new security vulnerabilities. The centralization of control and the complexity of virtualized infrastructures expand the attack surface, making traditional security mechanisms insufficient. This thesis proposes an AI-driven framework for intelligent intrusion detection in SDN/NFV environments. It addresses key limitations of existing approaches, including dataset inadequacy, class imbalance, and lack of real-time adaptability. The first contribution consists in a systematic benchmarking of machine learning and deep learning models on SDN-oriented datasets. The second contribution introduces SDN-Net, a unified and realistic dataset designed to improve detection performance across diverse attack scenarios. The third contribution explores reinforcement learning to enable adaptive and real-time intrusion detection and mitigation within the SDN control plane. Experimental results demonstrate the effectiveness of the proposed approach, achieving high detection performance in both binary and multi-class settings. This work highlights the potential of artificial intelligence to enhance network security and paves the way toward intelligent, adaptive, and autonomous defence mechanisms for next-generation networks.

Keywords

Software-Defined Networking (SDN), Network Function Virtualization (NFV), Intrusion Detection, Artificial Intelligence, Network Security, Dataset Fusion.