Multi-Armed Bandit-Based Client Scheduling for Federated Learning

By exploiting the computing power and local data of distributed clients, federated learning (FL) features ubiquitous properties such as reduction of communication overhead and preserving data privacy. In each communication round of FL, the clients update local models based on their own data and uplo...

Full description

Saved in:
Bibliographic Details
Published in:IEEE transactions on wireless communications Vol. 19; no. 11; pp. 7108 - 7123
Main Authors: Xia, Wenchao, Quek, Tony Q. S., Guo, Kun, Wen, Wanli, Yang, Howard H., Zhu, Hongbo
Format: Journal Article
Language:English
Published: New York IEEE 01.11.2020
The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
Subjects:
Algorithms
client scheduling
client selection
Clients
Communication
Computational modeling
Confidence
Convergence
Data models
Datasets
Federated learning
multi-armed bandit (MAB)
Multi-armed bandit problems
Scheduling
Smart phones
Training
Upper bound
Upper bounds
Wireless communication
ISSN:1536-1276, 1558-2248
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:By exploiting the computing power and local data of distributed clients, federated learning (FL) features ubiquitous properties such as reduction of communication overhead and preserving data privacy. In each communication round of FL, the clients update local models based on their own data and upload their local updates via wireless channels. However, latency caused by hundreds to thousands of communication rounds remains a bottleneck in FL. To minimize the training latency, this work provides a multi-armed bandit-based framework for online client scheduling (CS) in FL without knowing wireless channel state information and statistical characteristics of clients. Firstly, we propose a CS algorithm based on the upper confidence bound policy (CS-UCB) for ideal scenarios where local datasets of clients are independent and identically distributed (i.i.d.) and balanced. An upper bound of the expected performance regret of the proposed CS-UCB algorithm is provided, which indicates that the regret grows logarithmically over communication rounds. Then, to address non-ideal scenarios with non-i.i.d. and unbalanced properties of local datasets and varying availability of clients, we further propose a CS algorithm based on the UCB policy and virtual queue technique (CS-UCB-Q). An upper bound is also derived, which shows that the expected performance regret of the proposed CS-UCB-Q algorithm can have a sub-linear growth over communication rounds under certain conditions. Besides, the convergence performance of FL training is also analyzed. Finally, simulation results validate the efficiency of the proposed algorithms.
Bibliography:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
ISSN:1536-1276
1558-2248
DOI:10.1109/TWC.2020.3008091