Enabling Weak Client Participation via On-device Knowledge Distillation in Heterogeneous Federated Learning

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Lim, Jihyun, Jo, Junhyuk, Zhang, Tuo, Lee, Sunwoo
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912810024304640
author Lim, Jihyun
Jo, Junhyuk
Zhang, Tuo
Lee, Sunwoo
author_facet Lim, Jihyun
Jo, Junhyuk
Zhang, Tuo
Lee, Sunwoo
contents Online Knowledge Distillation (KD) is recently highlighted to train large models in Federated Learning (FL) environments. Many existing studies adopt the logit ensemble method to perform KD on the server side. However, they often assume that unlabeled data collected at the edge is centralized on the server. Moreover, the logit ensemble method personalizes local models, which can degrade the quality of soft targets, especially when data is highly non-IID. To address these critical limitations,we propose a novel on-device KD-based heterogeneous FL method. Our approach leverages a small auxiliary model to learn from labeled local data. Subsequently, a subset of clients with strong system resources transfers knowledge to a large model through on-device KD using their unlabeled data. Our extensive experiments demonstrate that our on-device KD-based heterogeneous FL method effectively utilizes the system resources of all edge devices as well as the unlabeled data, resulting in higher accuracy compared to SOTA KD-based FL methods.
format Preprint
id arxiv_https___arxiv_org_abs_2503_11151
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Enabling Weak Client Participation via On-device Knowledge Distillation in Heterogeneous Federated Learning
Lim, Jihyun
Jo, Junhyuk
Zhang, Tuo
Lee, Sunwoo
Machine Learning
Online Knowledge Distillation (KD) is recently highlighted to train large models in Federated Learning (FL) environments. Many existing studies adopt the logit ensemble method to perform KD on the server side. However, they often assume that unlabeled data collected at the edge is centralized on the server. Moreover, the logit ensemble method personalizes local models, which can degrade the quality of soft targets, especially when data is highly non-IID. To address these critical limitations,we propose a novel on-device KD-based heterogeneous FL method. Our approach leverages a small auxiliary model to learn from labeled local data. Subsequently, a subset of clients with strong system resources transfers knowledge to a large model through on-device KD using their unlabeled data. Our extensive experiments demonstrate that our on-device KD-based heterogeneous FL method effectively utilizes the system resources of all edge devices as well as the unlabeled data, resulting in higher accuracy compared to SOTA KD-based FL methods.
title Enabling Weak Client Participation via On-device Knowledge Distillation in Heterogeneous Federated Learning
topic Machine Learning
url https://arxiv.org/abs/2503.11151