ADVANCEMENTS IN AI

Prof. Vo Nguyen Quoc Bao

Deputy Vice Chancellor of Van Lang University, Vietnam

Bio: Vo Nguyen Quoc Bao (SMIEEE) is a Professor of Wireless Communications at Van Lang University, Vietnam. His research interests focus on advanced physical-layer techniques for 5G/6G systems, intelligent wireless networks, and AI-driven signal processing. He received his B.Eng. and M.Eng. degrees from Ho Chi Minh City University of Technology in 2002 and 2005, and his Ph.D. in Electrical Engineering from the University of Ulsan, South Korea, in 2010. Before joining Van Lang University in 2024, he spent over a decade at the Posts and Telecommunications Institute of Technology, where he was promoted to Full Professor in 2023.

Talk Title: Introduction to Van Lang University, Vietnam

Marie-Louise Cairns

Head of Regional Engagement for Northern Ireland at EPSRC

Talk Title: To be confirmed

Professor Pei Xiao

Institute for Communication Systems (ICS), Home for 5GIC & 6GIC, University of Surrey, Guildford, Surrey, GU2 7XH, United Kingdom.

Bio: Pei Xiao is a professor of Wireless Communications at the Institute for Communication Systems, home of 5GIC and 6GIC at the University of Surrey. He is the technical manager of 5GIC/6GIC, leading the research team in the new physical layer work area, and coordinating/supervising research activities across all the work areas (https://www.surrey.ac.uk/institute-communication-systems/5g-6g-innovation-centre). He has published extensively in the fields of communication theory, RF and antenna design, signal processing for wireless communications. Professor Xiao’s research focuses on next-generation wireless technologies, with a specific emphasis on IoT applications, satellite communications, and high-mobility vehicular networks. Key aspects of his work involve energy efficient cell-free network design, advanced multiple access technologies, new ISAC waveforms, and the application of artificial intelligence to wireless systems, including federated multi-agent reinforcement learning, graph neural network-based distributed optimisation, and collaborative edge inference. His contributions have significantly reduced hardware complexity and enhanced radio resource utilisation as well as power consumption through advanced signal processing and distributed learning techniques.

Talk Title: Distributed Federated Learning and Collaborative Inference for Next Generation Wireless Communications

Abstract: In this talk, we first discuss GNN-enhanced federated multi-agent reinforcement learning for energy-efficient distributed UAV communications. As UAV-assisted networks scale toward next-generation wireless systems, centralised training becomes untenable due to prohibitive backhaul overhead, privacy concerns, and single points of failure. To address this problem, we propose a federated GNN-enhanced MAPPO framework, where each UAV trains a local actor-critic model using neighbourhood observations only, with inter-agent reasoning preserved through a locally executed GNN on dynamic subgraphs. FedProx aggregation mitigates local drift from spatially non-IID traffic distributions, while transformer-based traffic prediction informs both UAV decision-making and federation round timing. Results show that FedMARL with GNN achieves robust multi-UAV cooperation and competitive energy efficiency without any centralised coordination, offering a scalable, privacy-preserving path to green UAV networks. In the second part of the talk, we introduce a novel collaborative inference architecture integrating cloud based LLMs with edge deployed small language models (SLMs). By enabling hierarchical KV-cache sharing and pipeline parallel processing, this framework significantly reduces inference latency and improves stability across diverse workload conditions.

Dr David Crawford

Principal Knowledge Exchange Fellow, University of Strathclyde

Bio: Dr Crawford is a Principal Knowledge Exchange Fellow at the University of Strathclyde, where he manages a programme of industry/academic collaborative projects focusing on shared spectrum, private 5G networks, and future advanced communications. David has championed the development of shared spectrum methods and adoption for more than a decade. David holds four degrees, all from Strathclyde, including a PhD and MBA. Prior to rejoining the University, initially to manage the Centre for White Space Communications, David worked in industry for Epson and Motorola.

Talk Title: To be confirmed

Dr Huy T. Nguyen

Assistant Professor, Van Lang University, Vietnam

Bio: Huy T. Nguyen (Member, IEEE) received his B.Eng. (Hons.) in Computer Science and Engineering from HCMC University of Technology, Vietnam, in 2013. He then earned his M.Eng. and Ph.D. degrees from the Department of Information and Communication System, Inje University, South Korea, in 2016 and 2019, respectively. He is currently a Lecturer (Assistant Professor) at the Faculty of Information Technology, Van Lang University (VLU), where he also leads the Smart and Autonomous Systems (SAS) Research Group. Before joining VLU, he served as Chief Technology Officer (CTO-Asia Desk) at Data Design Engineering (DDE), primarily developing data-driven AI solutions. His earlier career included significant contributions to the communication aspects of wireless SmartGrid systems as a Scientist at the Institute for Infocomm Research (I2R-A*STAR) and Nanyang Technological University (NTU), Singapore. His research interests lie in applying AI/ML and optimization techniques to signal processing and wireless communications for 5G and beyond.

Talk Title: Building Sustainable Connectivity: Energy Efficiency as a Pillar of Resilient 6G NTN

Dr. Viet Quoc Pham

Assistant Professor, Trinity College Dublin

Bio: Viet Quoc Pham is currently an Assistant Professor (above the bar) in Networks and Distributed Systems at the School of Computer Science and Statistics, Trinity College Dublin, Ireland. He earned his PhD degree (with Best PhD Dissertation Award) in Telecommunications Engineering from Inje University in 2017. He has special research interests in the areas of wireless AI, edge computing, Internet of Things, and distributed machine learning. He was a recipient of Clarivate Highly Cited Researcher Award in 2024 and IEEE ComSoc Best Young Researcher Award for EMEA 2023 in recognition of his research activities for the benefit of the Society. He is currently serving as an Editor of IEEE Communications Letters, IEEE Communications Standards Magazine, IEEE Communications Surveys & Tutorials, IEEE Transactions on Mobile Computing, and IEEE Transactions on Network Science and Engineering.

Talk Title: Rethinking Forgetting: Advances in Federated Machine Unlearning

Abstract: Federated machine unlearning (FUL) has emerged as a transformative paradigm for privacy-preserving and decentralized AI in 6G networks and the future Internet. FUL allows the removal of the data influence from a model trained across distributed clients, thereby upholding the right to be forgotten as mandated by privacy regulations. The talk will present the fundamental principles of federated unlearning and highlight recent advances through our research on verifiable federated unlearning and federated on-server unlearning.