EUVIMEDEUVIMED entdecken
Crossref

Distributed Large Models Training Optimization With Real-Time Wireless Channel Feedback

Jiaming Pei, Valerio Frascolla, Anwer Al-Dulaimi, Wei Liu, Theyazn H. H. Aldhyani, Ali Kashif Bashir, Shahid Mumtaz

IEEE Journal on Selected Areas in Communications · 2026 · 44 · 2231-2243

DOI: 10.1109/jsac.2025.3640136 ↗

Kurz gefasst

Fachartikel in der Zeitschrift IEEE Journal on Selected Areas in Communications, 2026. Verfasst von Jiaming Pei und 6 weiteren Autor:innen. Die Arbeit wurde bisher 49-mal zitiert (Crossref) und nennt 37 Literaturangaben. Gefördert von: National Natural Science Foundation of China, Dubai Future Foundation (DFF), Dubai, United Arab Emirates, Deanship of Scientific Research, Vice Presidency for Graduate Studies and Scientific Research, King Faisal University, Saudi Arabia.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Large-scale deep learning models rely on wireless networks for distributed training approaches, which are essential to meet the immense computational and data demands. However, the stochastic nature of wireless environments introduces significant challenges such as variable delays, noise interference, and packet loss, which lead to degraded gradient synchronization and hinder model convergence. In this work, we propose a novel communication-aware distributed training (CADT) framework that integrates real-time channel state information (CSI) feedback into the gradient aggregation process. Unlike conventional methods that assume static or ideal communication conditions, CADT dynamically reweights gradients from each node based on instantaneous channel quality, enabling robust aggregation under adverse wireless conditions. By dynamically adjusting the contribution of each node based on instantaneous channel conditions, CADT effectively compensates for wireless impairments, thereby ensuring more reliable gradient aggregation and significantly improving both convergence speed and final model accuracy. Extensive experiments on CIFAR-10, CIFAR-100, ImageNet, and SVHN using Vision Transformer and ResNet-50 demonstrate that CADT outperforms baseline methods in terms of convergence, accuracy, and communication efficiency. In addition, we provide a rigorous theoretical analysis that establishes convergence guarantees under realistic wireless conditions, thereby advancing the theoretical foundation of distributed optimization in non-ideal communication environments.Our framework offers a practical solution for real-world scenarios such as edge computing, where communication constraints and environmental variability are dominant factors.

Förderung

  • National Natural Science Foundation of China
  • Dubai Future Foundation (DFF), Dubai, United Arab Emirates
  • Deanship of Scientific Research, Vice Presidency for Graduate Studies and Scientific Research, King Faisal University, Saudi Arabia

Zitieren

APA

Pei, J., Frascolla, V., Al-Dulaimi, A., Liu, W., Aldhyani, T. H. H., Bashir, A. K., & Mumtaz, S. (2026). Distributed Large Models Training Optimization With Real-Time Wireless Channel Feedback. IEEE Journal on Selected Areas in Communications, 44, 2231-2243. https://doi.org/10.1109/jsac.2025.3640136

Vancouver

Pei J, Frascolla V, Al-Dulaimi A, Liu W, Aldhyani THH, Bashir AK, et al. Distributed Large Models Training Optimization With Real-Time Wireless Channel Feedback. IEEE Journal on Selected Areas in Communications. 2026;44:2231-2243. doi:10.1109/jsac.2025.3640136

BibTeX
@article{Pei2026,
  author = {Pei, Jiaming and Frascolla, Valerio and Al-Dulaimi, Anwer and Liu, Wei and Aldhyani, Theyazn H. H. and Bashir, Ali Kashif and Mumtaz, Shahid},
  title = {Distributed Large Models Training Optimization With Real-Time Wireless Channel Feedback},
  journal = {IEEE Journal on Selected Areas in Communications},
  year = {2026},
  volume = {44},
  pages = {2231--2243},
  doi = {10.1109/jsac.2025.3640136}
}
Kostenlos starten →