• Home
  • Search
  • An Efficient On-Device Federated Learning System Through the Interplay of Client Selection and Batch Size With Watermarked Data
  • Cite Icon1
  • https://doi.org/10.1109/tmc.2025.3585033Copy DOI Icon

An Efficient On-Device Federated Learning System Through the Interplay of Client Selection and Batch Size With Watermarked Data

Show More
  • Abstract
  • Literature Map
  • References
  • Citations
  • Similar Papers
Abstract

Federated Learning (FL) enables edge devices to collaboratively train a global model using local data. However, the increasing prevalence of watermarks in datasets presents a new challenge to efficient FL. While watermarks assert data ownership and copyright, they introduce complexities that can lead to shortcut learning problems and mislead utility measurements for client selection. These issues are further exacerbated by batch size variations in efficient FL frameworks, ultimately undermining their time-to-accuracy performance. We introduce <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">LotusFL</i>, an FL system designed to address the challenges posed by watermarked datasets in efficient FL. Specifically, it tackles the increased time-to-accuracy due to erroneous client selection and the accuracy degradation observed with larger batch sizes. <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">LotusFL</i> first estimates the characteristics of watermarks through statistical estimation and then adjusts the batch size using this estimated watermark information to balance the negative impact of the watermark against device idle waiting time. Additionally, its client selection mechanism, based on historical information, avoids the misleading utility signals from watermarks. This mechanism, working in conjunction with batch size adjustment, aims to accurately predict device runtime and identify potentially valuable devices. We evaluated <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">LotusFL</i> through a real-world deployment on 40 edge devices. Compared to state-of-the-art efficient FL frameworks, <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">LotusFL</i> achieves superior performance, enhancing accuracy by up to 8.2% and reducing training time by 1.97×.

Similar Papers
  • Conference Article
  • Citations6

Ed-Fed: A generic federated learning framework with resource-aware client selection for edge devices

  • Jun 18, 2023
  • Zitha Sasindran +2
  • Research Article
  • Citations3

SARS: A Personalized Federated Learning Framework Towards Fairness and Robustness against Backdoor Attacks

  • Nov 21, 2024
  • Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies
  • Weibin Zhang +4
  • Research Article
  • Citations11

Enabling efficient and low-effort decentralized federated learning with the EdgeFL framework

  • Oct 17, 2024
  • Information and Software Technology
  • Hongyi Zhang +2
  • Research Article
  • Citations19

FedDCS: A distributed client selection framework for cross device federated learning

  • Jul 01, 2023
  • Future Generation Computer Systems
  • Monalisa Panigrahi +2
  • Research Article
  • Citations51

Scalable and Low-Latency Federated Learning With Cooperative Mobile Edge Networking

  • Jan 01, 2024
  • IEEE Transactions on Mobile Computing
  • Zhenxiao Zhang +3
  • Research Article
  • Citations2

Efficient and Intelligent Multijob Federated Learning in Wireless Networks

  • Apr 01, 2025
  • IEEE Internet of Things Journal
  • Jiajin Wang +3
  • Research Article
  • Citations61

Optimizing Training Efficiency and Cost of Hierarchical Federated Learning in Heterogeneous Mobile-Edge Cloud Computing

  • May 01, 2023
  • IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems
  • Yangguang Cui +3
  • Research Article
  • Citations4

Joint Data Retransmission and Client Selection Optimization for Error-Tolerant Federated Learning in AAV Networks

  • Aug 01, 2025
  • IEEE Transactions on Cognitive Communications and Networking
  • Chao Dong +7
  • Research Article
  • Citations22

Blockchain-Based Self-Sovereign Identity: Taking Control of Identity in Federated Learning

  • Jan 01, 2024
  • IEEE Open Journal of the Communications Society
  • Engin Zeydan +6
  • Research Article
  • Citations90

Auction-Based Cluster Federated Learning in Mobile Edge Computing Systems

  • Apr 01, 2023
  • IEEE Transactions on Parallel and Distributed Systems
  • Renhao Lu +6
  • Research Article
  • Citations23

Wireless Federated Learning With Hybrid Local and Centralized Training: A Latency Minimization Design

  • Jan 01, 2023
  • IEEE Journal of Selected Topics in Signal Processing
  • Ning Huang +4
  • Conference Article
  • Citations2

Heterogeneous Federated Learning for Balancing Job Completion Time and Model Accuracy

  • Jan 01, 2023
  • Ruiting Zhou +4
  • Research Article
  • Citations12

Two-Stage Client Selection for Federated Learning Against Free-Riding Attack: A Multiarmed Bandits and Auction-Based Approach

  • Oct 15, 2024
  • IEEE Internet of Things Journal
  • Renhao Lu +9
  • Research Article
  • Citations45

Hierarchical Clustering-based Personalized Federated Learning for Robust and Fair Human Activity Recognition

  • Mar 27, 2023
  • Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies
  • Youpeng Li +2
  • PDF
  • Research Article
  • Citations31

Comparative analysis of open-source federated learning frameworks - a literature-based survey and review

  • Jun 28, 2024
  • International Journal of Machine Learning and Cybernetics
  • Pascal Riedel +5
Cactus Communications logo

Copyright 2026 Cactus Communications. All rights reserved.