IEEE Québec Section Webinar: Resilient Transceiver Architectures for EMI-Challenged Smart Grid Communications

[]

Zoom Link: <a href="https://ulaval.zoom.us/j/69550214937?pwd=iH5u8TmnfzeZUxEc4LMpU1IupTAwvh.1

Time:” target=”_blank” title=”https://ulaval.zoom.us/j/69550214937?pwd=iH5u8TmnfzeZUxEc4LMpU1IupTAwvh.1

Time:”>https://ulaval.zoom.us/j/69550214937?pwd=iH5u8TmnfzeZUxEc4LMpU1IupTAwvh.1

Time: 12 December 2025, 11.00 AM EST to 12.00 PM EST

Talk Abstract:

The reliable operation of smart grids increasingly relies on wireless communication links deployed within high-voltage substations and distribution infrastructures. However, these environments are dominated by severe electromagnetic interference (EMI), producing bursty, high-amplitude impulsive noise with strong temporal correlation. Conventional transceiver design based on simple clipping/blanking, or memoryless soft decoding fails to ensure reliable connectivity under realistic EMI, resulting in critical degradation of QoS. This talk presents promising EMI-aware transceiver architectures that bridge theoretical modeling and practical resilience. We first revisit EMI characterization in smart grids, highlighting the impulsive, bursty, and dynamic nature of EMI. We then explore transceiver design strategies ranging from enhanced LLR-based detection to AI-driven architectures. Finally, we present fully AI-native deep semantic transceivers that jointly optimize encoding, decoding, and noise mitigation, demonstrating robust communication in presence of strong <a href="http://EMI.

Speaker” target=”_blank” title=”EMI.

Speaker”>EMI.

Speaker Biography:

Georges Kaddoum is a professor and the research director of the Resilient Machine Learning Institute (ReMI) at École de Technologie Supérieure (ÉTS), Université du Québec, Montréal, Canada. He also holds an industrial research chair and a Tier 2 Canada Research Chair. He earned his Ph.D. in Signal Processing and Telecommunications with High Honors from the National Institute of Applied Sciences (INSA), University of Toulouse, France, in 2009. His research focuses on wireless communication networks, tactical communications, resource allocation, and network security. Prof. Kaddoum is a member of the Royal Society of Canada and has received multiple prestigious recognitions. He has served as an associate editor for IEEE Transactions on Information Forensics and Security and IEEE Communications Letters. Currently, he is an area editor for IEEE Transactions on Machine Learning in Communications and Networking and an editor for IEEE Transactions on <a href="http://Communications.

Meeting” target=”_blank” title=”Communications.

Meeting”>Communications.

Meeting Link: https://ulaval.zoom.us/j/69550214937?pwd=iH5u8TmnfzeZUxEc4LMpU1IupTAwvh.1, Quebec City, Quebec, Canada