Unlocking the Power of Large Language Models in Wireless Networks: From Prompt Engineering to Intelligent Optimization
Large Language Models (LLMs) are reshaping wireless networks through their powerful reasoning and adaptability. This talk begins with LLM fundamentals and the unique challenges of applying them in wireless systems, such as high computational demands, limited datasets, and real-time constraints. Prompt engineering will be introduced as a lightweight alternative to fine-tuning, enabling accurate, context-aware outputs under device limitations. Strategies including in-context learning, chain-of-thought prompting, compression, and self-refinement will be discussed, along with novel iterative and self-refined prompting techniques that can match or surpass traditional machine learning methods while reducing complexity. We will then examine three key wireless network use cases: (i) Network Resource Allocation: iterative prompting for near-optimal optimization without retraining; (ii) Prediction: a self-refined LLM that improves wireless traffic forecasting accuracy by over 20% in dynamic environments; and (iii) Decision-Making for Autonomous Systems: a hybrid LLM–Double Deep Q-Network framework that jointly optimizes vehicle-to-infrastructure (V2I) communications and autonomous driving policies. Case studies demonstrate faster convergence, greater adaptability, and superior performance over conventional methods, positioning LLMs as a unifying intelligence layer for next-generation AI-driven networks.
Date and Time
Location
Hosts
Registration
-
Add Event to Calendar
- 6 Rue Richard Coudenhove-Kalergi, 1359 Kirchberg Luxembourg
- Luxembourg, Luxembourg
- Luxembourg 1359
- Building: Batiment Central
- Room Number: SALLE DES CONSEILS D17
Speakers
Hina of York university, Toronto, Canada
Unlocking the Power of Large Language Models in Wireless Networks: From Prompt Engineering to Intelligent Optimization
Large Language Models (LLMs) are reshaping wireless networks through their powerful reasoning and adaptability. This talk begins with LLM fundamentals and the unique challenges of applying them in wireless systems, such as high computational demands, limited datasets, and real-time constraints. Prompt engineering will be introduced as a lightweight alternative to fine-tuning, enabling accurate, context-aware outputs under device limitations. Strategies including in-context learning, chain-of-thought prompting, compression, and self-refinement will be discussed, along with novel iterative and self-refined prompting techniques that can match or surpass traditional machine learning methods while reducing complexity. We will then examine three key wireless network use cases: (i) Network Resource Allocation: iterative prompting for near-optimal optimization without retraining; (ii) Prediction: a self-refined LLM that improves wireless traffic forecasting accuracy by over 20% in dynamic environments; and (iii) Decision-Making for Autonomous Systems: a hybrid LLM–Double Deep Q-Network framework that jointly optimizes vehicle-to-infrastructure (V2I) communications and autonomous driving policies. Case studies demonstrate faster convergence, greater adaptability, and superior performance over conventional methods, positioning LLMs as a unifying intelligence layer for next-generation AI-driven networks.
Biography:
Hina Tabassum (Senior Member, IEEE) received the Ph.D. degree from the King Abdullah University of Science and Technology, Thuwal, Saudi Arabia. She is currently an Associate Professor with the Lassonde School of Engineering, York University, Toronto, where she joined as an Assistant Professor in 2018. She was appointed as a Visiting Faculty with University of Toronto, Toronto, ON, Canada, in 2024, and the York Research Chair of 5G/6G-enabled mobility and sensing applications in 2023, for five years. She has coauthored more than 120 refereed articles in well-reputed IEEE journals, magazines, and conferences. Her current research interests include multi-band 6G wireless communications and sensing networks, connected and autonomous systems, AI-enabled network mobility, and resource management solutions. She has been selected as the IEEE ComSoc Distinguished Lecturer for the term 2025–2026. She is listed in Stanford’s list of the World’s Top Two-Percent Researchers in 2021–2024. She was the recipient of the Lassonde Innovation Early-Career Researcher Award in 2023 and the N2Women: Rising Stars in Computer Networking and Communications in 2022. She was the Founding Chair of the Special Interest Group on THz communications in the IEEE Communications Society–Radio Communications Committee. Currently, she is serving as an Area Editor of the IEEE Open Journal of the Communications Society and IEEE Communications Surveys and Tutorials as well as an Associate Editor for IEEE Transactions on Communications, IEEE Transactions on Mobile Computing, and IEEE Transactions on Wireless Communications.
Email:
Address:Lassonde School of Engineering, Electrical Engineering and Computer Science, Toronto, Ontario, Canada