1st International Workshop on Federated Learning for Wireless Edge Artificial Intelligence (FedEdgeAI)

In conjunction with the 45th IEEE International Conference on Distributed Computing Systems (IEEE ICDCS 2025)


20 July - 23 July, 2025, Glasgow, Scotland, UK

About the Workshop

Welcome to the 1st workshop on federated learning for Wireless Edge AI (FedEdgeAI). FedEdgeAI will be hosted in conjunction with the IEEE ICDCS 2025 conference, which will be held in Glasgow, Scotland, United Kingdom, from July 20th to July 23rd, 2025.

We invite you to submit your original work on topics related to federated learning, focusing on real-world challenges when federated learning is deployed in practical scenarios. This includes algorithms for distributed machine learning, adaptive techniques for changing network conditions, edge AI resilience, benchmarking generative models at the edge, downsizing Large Language Models (LLMs) into Small Language Models (SLMs) for improved computation and communication efficiency, semantic communication, asynchronous federated learning training, and rethinking communication protocols for wireless federated learning.

Call for Papers

Edge AI emerged as an evolution of the edge computing paradigm, deploying AI algorithms and models directly on edge devices. Within this context, the concept of federated learning provides privacy by design in an machine learning technique, enabling collaborative learning across multiple distributed devices without sending raw data to a central server while processing data locally on devices. However, given the limited availability of resources on many devices, performing federated learning on such devices is impractical due to increased training times. Moreover, for training machine learning models that may be a Deep Neural Network (DNN), massive amounts of parameter updates need to be synchronized across distributed devices, creating potential congestion and eventually slowdowns the entire training process.

Specifically, the end devices used in federated learning are predominantly wireless and typically operate with limited bandwidth, such as 2G, 3G, or Wi-Fi. The global and local model parameters use uplink and downlink transmission, which depend on bandwidth resource block allocation, fading, and interference from others. Exchanging model parameters over such lossy networks may result in challenges such as transmission delay, which impacts the convergence time of the model, and packet losses, which affect the model's accuracy.

We invite submissions on a wide range of topics including, but not limited to:

Important Dates

  • Paper Submission Deadline:
  • Notification of Acceptance: 2 April 2025
  • Workshop Papers Camera-Ready: 16 April 2025 (this is a hard deadline)
  • Workshop Date:  20 July 2025
  • *All deadlines are Anywhere on Earth (AoE) .

Submission Guidelines

Authors are invited to submit original and unpublished work, which must not be submitted concurrently for publication elsewhere, in the following format:

Organizing Committee

General Chairs

Dr. Rehmat Ullah

Dr. Rehmat Ullah

Senior Lecturer (Associate Professor) at Newcastle University, UK

Dr. John Doe

Dr. Danh Le-Phuoc

Head of PICOM Lab and DFG Principle Investigator, Technische Universität Berlin, Germany.

Dr. John Doe

Dr. Muhammad Atif Ur Rehman

Lecturer (Assistant Professor) at Manchester Metropolitan University, UK.

Dr. John Doe

Dr. Ahmed M. A. Sayed

Senior Lecturer (Associate Professor) and Head of SAYED Systems Group at Queen Mary University of London, UK.

Publicity Chair

Dr. Kevin Li

Dr. Kevin Li

Post-doc researcher, Queen Mary University of London, UK.

Program Committee Members

Keynote Speakers

Dr. José Cano

Dr. José Cano

Associate Professor in School of Computing Science at University of Glasgow, UK.

Talk Title: Accelerating LLMs at the Edge: The Power of Efficient HW/SW Co-Design

Dr. Edoardo Prezioso

Dr. Edoardo Prezioso

Research Fellow at University of Naples Federico II, Italy.

Talk Title: Federated Learning under Operational Constraints: A Study on Statistical and Device Heterogeneity

Program

20 July 2025

Venue: Radisson Blu Hotel, Glasgow, Scotland, UK

13:45 – 13:55 Introduction (Workshop and Organizers Introduction)
13:55 – 14:40
Keynote 1: Accelerating LLMs at the Edge: The Power of Efficient HW/SW Co-Design
Speaker: Dr. José Cano, Associate Professor, School of Computing Science, University of Glasgow, UK

José Cano is an Associate Professor in the School of Computing Science at the University of Glasgow, where he leads the Glasgow Intelligent Computing Laboratory (gicLAB) within the Systems Research Section (GLASS) and is also deputy Head of GLASS. His research interests are in the broad areas of Computer Architecture, Computer Systems, Compilers, Machine Learning, and Security. His current research is mainly focused on Hardware/Software co-design approaches to efficiently and securely deploy AI/ML applications on resource-constrained edge devices.

José is currently Principal Investigator at the University of Glasgow on the EU’s Horizon Europe project dAIEDGE and the UKRI APRIL AI Hub project SECDA-DSE, and Co-Investigator on the UKRI project IDEAL . He was Principal Investigator on the UK’s PETRAS project MAISE , and Co-Investigator on the UKRI "Digital Security by Design" projects AppControl and Morello-HAT . He has obtained over £550K as a project PI, over £1.3M as a project Co-I, and over £81K from eight personal grants.

José received his Ph.D. in Computer Science from Universitat Politècnica de València (Spain) in January 2012. After that, he was a Postdoctoral Researcher in the Department of Computer Architecture at Universitat Politècnica de Catalunya (Spain) until December 2013. Then he joined the Institute for Computing Systems Architecture in the School of Informatics at The University of Edinburgh (UK) as a Research Associate between January 2014 and August 2018. He is a senior member of the IEEE and ACM research societies and a member of the HiPEAC, dAIEDGE, and PETRAS networks of excellence.

Abstract:

Large Language Models (LLMs) are increasingly a key component of Artificial Intelligence (AI) applications across various domains, including computer vision, natural language processing, and scientific computing. Executing LLMs on edge devices can enable secure computation, lower energy consumption, and reduce costs. However, to be practical, their performance must improve significantly due to the demanding memory and compute requirements of emerging LLMs, which conflict with the limitations of resource-constrained edge devices.

In this talk, I will introduce the Glasgow Intelligent Computing Laboratory (gicLAB) and give an overview of our current and future research, with an emphasis on approaches to efficiently and securely deploy and run LLMs on resource-constrained edge devices.

Session 1: Federated IoT Environments (14:40 - 14:55)
FedHDC-IDS: A Hyperdimensional Computing Based Network Intrusion Detection System for IoT Networks Authors: Othmane Belarbi, Yogha Restu Pramadi, Omer Rana, Yuhua Li, Aftab Khan, Theodoros Spyridopoulos
A Multi-Criteria Selection of IoT Devices For Federated Learning with Non-IID Data Authors: Hajar El Hammouti, Mustafa Kishk, Ahmed Elzenaty
15:15 – 15:45 Coffee Break
15:45 – 16:15
Keynote 2: Federated Learning under Operational Constraints: A Study on Statistical and Device Heterogeneity
Speaker: Dr. Edoardo Prezioso, Senior Research Fellow at University of Naples Federico II, Italy. .

Edoardo Prezioso is currently a Senior Research Fellow within the Department of Mathematics and Applications "R. Caccioppoli" at the University of Naples Federico II, Italy. His research activities are primarily conducted as a core member of the M.O.D.A.L. (Mathematical mOdelling and Data AnaLysis) research group and laboratory, a dynamic, interdisciplinary unit focused on cutting-edge research in Artificial Intelligence.

His scientific productivity is demonstrated by a portfolio of over 30 publications, reflecting a significant impact in his field. He has been a speaker at numerous international conferences and is a co-inventor of a patent for multivariate time-series prediction.

Session 2: Federated Optimisation (16:15 - 16:45)
Decentralized Intelligence: A Digital Twin-Assisted Hierarchical Federated Deep Reinforcement Learning Framework for Resource Optimization Authors: Ahmad Arsalan, Wai Fong Tam, Tariq Umer, Rana Asif Rehman, Ahmed M. Abdelmoniem
Benchmarking Mutual Information-based Loss Functions in Federated Learning Authors: Sarang S, Harsh D Chothani, Qilei Li, Ahmed M. Abdelmoniem, Arnab Paul
Panel and Closing Remarks (16:45 - 17:15)

Contact

FedEdgeAI Chairs

Email: fededgeai@gmail.com

National Edge AI Hub

School of Computing at Newcastle University, UK