FedDCT: A Dynamic Cross-Tier Federated Learning Framework in Wireless Networks

Fuente: arXiv
Enregistré dans:
Détails bibliographiques
Auteurs principaux: Xian, Youquan, Gan, Xiaoyun, Yao, Chuanjian, Li, Dongcheng, Wang, Peng, Liu, Peng, Zhao, Ying
Format: Preprint
Publié: 2023
Sujets:
Accès en ligne:
Tags: Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
_version_ 1866913580652167168
author Xian, Youquan
Gan, Xiaoyun
Yao, Chuanjian
Li, Dongcheng
Wang, Peng
Liu, Peng
Zhao, Ying
author_facet Xian, Youquan
Gan, Xiaoyun
Yao, Chuanjian
Li, Dongcheng
Wang, Peng
Liu, Peng
Zhao, Ying
contents Federated Learning (FL), as a privacy-preserving machine learning paradigm, trains a global model across devices without exposing local data. However, resource heterogeneity and inevitable stragglers in wireless networks severely impact the efficiency and accuracy of FL training. In this paper, we propose a novel Dynamic Cross-Tier Federated Learning framework (FedDCT). Firstly, we design a dynamic tiering strategy that dynamically partitions devices into different tiers based on their response times and assigns specific timeout thresholds to each tier to reduce single-round training time. Then, we propose a cross-tier device selection algorithm that selects devices that respond quickly and are conducive to model convergence to improve convergence efficiency and accuracy. Experimental results demonstrate that the proposed approach under wireless networks outperforms the baseline approach, with an average reduction of 54.7\% in convergence time and an average improvement of 1.83\% in convergence accuracy.
format Preprint
id arxiv_https___arxiv_org_abs_2307_04420
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle FedDCT: A Dynamic Cross-Tier Federated Learning Framework in Wireless Networks
Xian, Youquan
Gan, Xiaoyun
Yao, Chuanjian
Li, Dongcheng
Wang, Peng
Liu, Peng
Zhao, Ying
Distributed, Parallel, and Cluster Computing
Artificial Intelligence
Federated Learning (FL), as a privacy-preserving machine learning paradigm, trains a global model across devices without exposing local data. However, resource heterogeneity and inevitable stragglers in wireless networks severely impact the efficiency and accuracy of FL training. In this paper, we propose a novel Dynamic Cross-Tier Federated Learning framework (FedDCT). Firstly, we design a dynamic tiering strategy that dynamically partitions devices into different tiers based on their response times and assigns specific timeout thresholds to each tier to reduce single-round training time. Then, we propose a cross-tier device selection algorithm that selects devices that respond quickly and are conducive to model convergence to improve convergence efficiency and accuracy. Experimental results demonstrate that the proposed approach under wireless networks outperforms the baseline approach, with an average reduction of 54.7\% in convergence time and an average improvement of 1.83\% in convergence accuracy.
title FedDCT: A Dynamic Cross-Tier Federated Learning Framework in Wireless Networks
topic Distributed, Parallel, and Cluster Computing
Artificial Intelligence
url https://arxiv.org/abs/2307.04420