Program

Location: Jacques Morgenstern Amphitheatre, Kahn building, Inria Centre at Université Côte d’Azur, 2004 Route des Lucioles, 06410 Biot, France
Time Tuesday 8 Sep Wednesday 9 Sep Thursday 10 Sep Friday 11 Sep
08:30-09:00 Registration
09:00-10:30 Opening Keynote: AI in radio access networks: challenges and open problems
Speaker: Zwi Altman (Orange Lab)
Course: Robust Machine Learning: A Quest to Learning in Untrusted Environment (I)
Speaker: Nirupam Gupta (University of Copenhagen)
Course: Robust Machine Learning: A Quest to Learning in Untrusted Environment (II)
Speaker: Nirupam Gupta (University of Copenhagen)
Keynote: Distributed AI at the Edge
Speaker: Alexandru Dobrila (Hivenet/Antimatter)
10:30-11:00 Break
11:00-12:30 Lecture: Distributed inference in the edge-network-cloud continuum
Speaker: Frédéric Giroire (CNRS)
Lecture: Bayesian optimization: Theory and applications to telecommunications
Speaker: Lorenzo Maggi (NVIDIA)
Course: Reinforcement learning: from bandits to structured MDPs (I)
Speaker: Bruno Gaujal (Inria)
Course: Reinforcement learning: from bandits to structured MDPs (II)
Speaker: Bruno Gaujal (Inria)
12:30-14:00 Lunch
14:00-15:30 Lecture: The Road to Autonomous Networks: AI/ML Foundations and Frontiers
Speaker: Farnaz Moradi (Ericsson)
Panel: Application and Role of AI in Industry
Panelists:
  • Zwi Altman (Orange)
  • Alexandru Dobrila (Hivenet/Antimatter)
  • Lorenzo Maggi (NVIDIA)
  • Farnaz Moradi (Ericsson)
  • Student Presentations (Session 3) Student Presentations (Session 5)
    15:30-16:00 Break END
    16:00-17:30 Student Presentations (Session 1) Student Presentations
    (Session 2)
    Student Presentations (Session 4)
    19:00-21:30 Social Dinner at l’Atelier 67


    Session 1 (Tuesday 8 Sep): Efficient Edge Inference

    Léo Bernard — Threshold-based Routing for Energy-efficient Experts
    Kyrylo Tymchenko — Confidence-Shaped Regression Cascades for Efficient Edge Inference

    Session 2 (Wednesday 9 Sep): Resource Allocation and Efficient Training

    Isidoor Pinillo Esquivel — Quantized Online Gradient Descent for Constant-Time Caching with Dynamic Regret Guarantees
    Mariya Peter — Fairness-Aware Optimal Transport Framework for Flow Allocation in Interconnected Network Systems
    Nicolas Helson — Network-Adaptive Gradient Compression for Faster ML Model Training in Datacenters

    Session 3 (Thursday 10 Sep): Federated Learning: Fairness, Robustness, and Unlearning

    Neeraja Sudhakaran — Towards Fair Federated Learning: Analyzing the Impact of Data Heterogeneity and Data Repair
    Jingye Wang — The Praetorian Guard: Unveiling the Vulnerability of Trust-Based Defense in Federated Learning
    Florian Zimmer — Enabling Decentralised Federated Unlearning in Industrial Cross-Organisational Environments
    Panos Raptis — Dynamic Fairness in Multi-Task Edge Learning

    Session 4 (Thursday 10 Sep): Reinforcement Learning and AI-Driven Network Intelligence

    Sooraj Skanda — Model-Based Multi-Agent Reinforcement Learning for Integrated Sensing and Communication
    Andreas Pattichis — Continual Learning from Streams of Unlabeled Data in LLM Systems
    Michele Simeone — Exploring KPI Trade-offs in O-RAN Conflict Resolution with Deep Reinforcement Learning
    Luigi Rachiele — From Fixed Pipelines to Agentic Decisions: An LLM Orchestrator for Anomaly Detection in Mobile Network Drive Tests

    Session 5 (Friday 11 Sep): Learning and Control under Resource Constraints

    Muyun Li — RL-based Edge Access Control for Time-Sensitive Tasks under Resource Contention
    Oihan Azkarate Iriarte — Optimal Control and Learning of Tandem Queues with Transfer Costs
    Haoming Lin — Structure of Optimal Admission Control under Resource Constraints
    Berfin Dinc — Closed-Loop Control with Delayed Information