When One-Shot Federated Learning Meets Diffusion Models at the Edge: Technological Advances and Applications
Journal of Intelligent Computing and Networking · 2026
DOI: 10.64509/jicn.21.64 ↗Kurz gefasst
Fachartikel in der Zeitschrift Journal of Intelligent Computing and Networking, 2026. Verfasst von Wanxiang Chen und 4 weiteren Autor:innen. Die Arbeit wurde bisher 2-mal zitiert (Crossref) und nennt 113 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Abstract (Originalsprache)
The rapid growth of the Internet of Things (IoT) and edge devices has accelerated the adoption of edge computing, which processes data at the edge to reduce latency and enhance privacy. In this context, federated learning (FL) has emerged as a promising framework for distributed model training without sharing raw data. However, traditional FL methods are often impractical in edge scenarios due to their reliance on extensive, resource-intensive communication rounds. To tackle this issue, one-shot federated learning (OSFL) has been proposed, enabling model aggregation in a single communication round. Meanwhile, diffusion models have gained significant attention for their powerful generative capabilities, especially in image synthesis and data augmentation. Recently, researchers have begun exploring the implementation of OSFL with diffusion models. By leveraging diffusion-based data generation, these approaches efficiently combine knowledge from distributed sources. This synergy not only improves model performance under non-IID data but also addresses the challenges related to data scarcity and privacy in edge environments. In this review, we systematically analyze the intersection of these two advanced paradigms, highlighting their complementarity and discussing key design considerations. Furthermore, we outline the contributions of this work: (1) providing a comprehensive taxonomy of existing approaches that combine OSFL with diffusion models; (2) identifying open challenges and future research directions; and (3) offering practical insights for deploying such integrated systems in real-world edge computing applications.
Zitieren
APA
Chen, W., Deng, D., Xiang, C., Liu, Z., & Xiao, B. (2026). When One-Shot Federated Learning Meets Diffusion Models at the Edge: Technological Advances and Applications. Journal of Intelligent Computing and Networking. https://doi.org/10.64509/jicn.21.64
Vancouver
Chen W, Deng D, Xiang C, Liu Z, Xiao B. When One-Shot Federated Learning Meets Diffusion Models at the Edge: Technological Advances and Applications. Journal of Intelligent Computing and Networking. 2026. doi:10.64509/jicn.21.64
BibTeX
@article{Chen2026,
author = {Chen, Wanxiang and Deng, Dongshang and Xiang, Chaocan and Liu, Zhi and Xiao, Bin},
title = {When One-Shot Federated Learning Meets Diffusion Models at the Edge: Technological Advances and Applications},
journal = {Journal of Intelligent Computing and Networking},
year = {2026},
doi = {10.64509/jicn.21.64}
}
EUVIMED