8th International Conference on Machine Learning for Networking (MLN'2025)
Keynote: Yassine Hadjadj Aoul, Professor at University of Rennes, France
Title: Graph Neural Networks for Scalable Network Optimization
Abstract
Modern networks face unprecedented complexity driven by traffic explosion, service diversification, and the integration of billions of heterogeneous devices. To address these challenges, network slicing and network tomography emerge as pivotal enablers for flexible resource provisioning and enhanced observability. This presentation introduces GNN-driven strategies to master two core challenges: network slicing and network tomography. By modelling networks as graphs, Graph Neural Networks (GNNs) surpass the limitations of traditional heuristics and even Deep Reinforcement Learning (DRL), delivering scalable, adaptive, and generalizable solutions for resource orchestration and performance inference. These advances enable robust service deployment under diverse QoS demands and enhance network observability through smart, data-driven insights. Ultimately, GNNs pave the way for autonomous, zero-touch network management in the 6G era and beyond.
Biography
Yassine Hadjadj-Aoul is a Full Professor at the University of Rennes, France, where he heads the ESIR’s school computer science department since 2022. He is affiliated with the INRIA Ermine team-project and the IRISA Laboratory, which he joined in 2009 when he was hired as an Associate Professor at the University of Rennes. After completing his Ph.D. from the University of Versailles, he further developed his research experience as a postdoctoral researcher at the University of Lille 1 and as a Marie Curie Fellow under the EU FP6 EIF program at University College Dublin (UCD), Ireland. A Senior Member of the IEEE and recipient of multiple best paper awards, he has supervised numerous Ph.D. students and postdoctoral researchers and has played a key role in several high-impact projects in the area of Next Generation Networks. He actively contributes to scientific committees and editorial boards, and co-leads both the "AI for Infrastructure" initiative within the French GDR and the SmartNet research action between INRIA and Nokia Bell Labs, which focuses on AI-driven intelligent network management. His primary research interests include Congestion Control, Network Slicing, and more generally on QoS/QoE provisioning across the Edge–Cloud continuum.