Tutorial

Too Hot, Too Small, Too Dynamic:
Thermal Behavior, Device Constraints, and Robust ML for XR Wearables

Date: September 16, 2026

XR is rapidly advancing toward the metaverse, but delivering high-quality experiences on lightweight wearables remains difficult due to limited processing, battery, and thermal capacity. While edge offloading helps support demanding tasks such as rendering and perception, it also introduces latency, bandwidth, and link reliability challenges. At the same time, increasing on-device computation improves responsiveness but raises power consumption and heat, potentially harming user comfort and QoE. In parallel, XR increasingly depends on AI/ML for tasks such as scene understanding and content generation, yet conventional models assume stationary data and struggle with the highly dynamic, non-stationary nature of XR environments. This tutorial explores the intersection of XR system design, thermal-aware computing, and time-resilient ML for robust immersive experiences.

Requirements for Attendees

This tutorial is intended for researchers, engineers, and students in computer science, telecommunications, networking, and human–computer interaction. It will also be of interest to participants working on wearable technologies, XR systems, network optimization, thermal management, machine learning, and, more broadly, the convergence of these areas. Attendees are not expected to bring any particular hardware or complete any preparation in advance.

Organizers

Olga Chukhno, Mediterranea University of Reggio Calabria

Olga Chukhno received her MSCA Innovative Training Network fellowship and obtained her double Ph.D. from Tampere University (Finland) and Mediterranea University of Reggio Calabria (Italy), where she is currently an Assistant Professor. Her main research interests include wireless communications, programmable heterogeneous networks, advanced algorithms for managing distributed services, and resource optimization with a particular focus on XR applications. She was recently recognized as a "Rising Star" in the "100 Brilliant and Inspiring Women in 6G" list as well as in computer networking and communications. She actively participates in research projects (currently, the RESTART program funded by the European Union; previously, the SNS JU ADROIT6G project). She is currently an Associate Editor for IEEE Communications Letters (recognized as Exemplary Editor in 2024 and 2025) and Review Editor at Frontiers. She served as a TPC member in IEEE GLOBECOM, IEEE ISC2, WiMob, IEEE ISCC, IEEE MEDITCOM, ACM (GoodIT, ICDCN), IEEE PerCom, and IEEE VTC2025-Spring. Since 2024, she has been teaching courses on metaverse and mathematical modeling and simulation for programmable networks for Master's and Ph.D. students.

Nadezhda Chukhno, IMDEA Networks Institute

Nadezhda Chukhno obtained her double Ph.D. from Mediterranea University of Reggio Calabria (Italy) and Jaume I University (Spain) within the European Union’s Horizon 2020 MSCA Innovative Training Network/European Joint Doctorate. She is currently the principal investigator of the MSCA Postdoctoral Fellowship 6G-AI-TANGO project at IMDEA Networks Institute (Spain), where she works on wireless communications, cellular traffic data analysis, drift detection, and the design of time-resilient AI models for mobile network forecasting, including large language models and few-shot learning. She has served as a reviewer for international journals such as IEEE COMML, IEEE Communications Standards Magazine, IEEE Network Magazine, IEEE TWC, Transactions on Emerging Telecommunications Technologies, IEEE TCOM, IEEE OJ-COMS, IEEE TVT, IEEE T-ITS, IEEE TBC, IEEE Sensors, Wireless Networks, and Computer Networks. She has also served as a TPC member and reviewer for IEEE INFOCOM, IEEE GLOBECOM, IEEE VTC, IEEE BMSB, IEEE CCNC, WiMob, IEEE ISCC, IEEE ISC2, and ICUMT. In 2024 and 2025, she was selected as an Exemplary Reviewer for IEEE COMML, and in 2024 she was nominated as a Rising Star in the “100 Brilliant and Inspiring Women in 6G” list.