TY - JOUR
T1 - Client Selection and Bandwidth Allocation in Wireless Federated Learning Networks
T2 - A Long-Term Perspective
AU - Xu, Jie
AU - Wang, Heqiang
N1 - Funding Information:
Manuscript received April 15, 2020; revised August 27, 2020; accepted October 9, 2020. Date of publication October 22, 2020; date of current version February 11, 2021. This work was supported in part by the NSF under Grant ECCS-2033681, Grant CNS-2006630, and Grant ECCS-2029858. The associate editor coordinating the review of this article and approving it for publication was K. Choi. (Corresponding author: Jie Xu.) The authors are with the Department of Electrical and Computer Engineering, University of Miami, Coral Gables, FL 33146 USA (e-mail: jiexu@miami.edu).
Publisher Copyright:
© 2002-2012 IEEE.
PY - 2021/2
Y1 - 2021/2
N2 - This paper studies federated learning (FL) in a classic wireless network, where learning clients share a common wireless link to a coordinating server to perform federated model training using their local data. In such wireless federated learning networks (WFLNs), optimizing the learning performance depends crucially on how clients are selected and how bandwidth is allocated among the selected clients in every learning round, as both radio and client energy resources are limited. While existing works have made some attempts to allocate the limited wireless resources to optimize FL, they focus on the problem in individual learning rounds, overlooking an inherent yet critical feature of federated learning. This paper brings a new long-term perspective to resource allocation in WFLNs, realizing that learning rounds are not only temporally interdependent but also have varying significance towards the final learning outcome. To this end, we first design data-driven experiments to show that different temporal client selection patterns lead to considerably different learning performance. With the obtained insights, we formulate a stochastic optimization problem for joint client selection and bandwidth allocation under long-term client energy constraints, and develop a new algorithm that utilizes only currently available wireless channel information but can achieve long-term performance guarantee. Experiments show that our algorithm results in the desired temporal client selection pattern, is adaptive to changing network environments and far outperforms benchmarks that ignore the long-term effect of FL.
AB - This paper studies federated learning (FL) in a classic wireless network, where learning clients share a common wireless link to a coordinating server to perform federated model training using their local data. In such wireless federated learning networks (WFLNs), optimizing the learning performance depends crucially on how clients are selected and how bandwidth is allocated among the selected clients in every learning round, as both radio and client energy resources are limited. While existing works have made some attempts to allocate the limited wireless resources to optimize FL, they focus on the problem in individual learning rounds, overlooking an inherent yet critical feature of federated learning. This paper brings a new long-term perspective to resource allocation in WFLNs, realizing that learning rounds are not only temporally interdependent but also have varying significance towards the final learning outcome. To this end, we first design data-driven experiments to show that different temporal client selection patterns lead to considerably different learning performance. With the obtained insights, we formulate a stochastic optimization problem for joint client selection and bandwidth allocation under long-term client energy constraints, and develop a new algorithm that utilizes only currently available wireless channel information but can achieve long-term performance guarantee. Experiments show that our algorithm results in the desired temporal client selection pattern, is adaptive to changing network environments and far outperforms benchmarks that ignore the long-term effect of FL.
KW - Federated learning (FL)
KW - client selection
KW - resource allocation
KW - wireless networks
UR - http://www.scopus.com/inward/record.url?scp=85101478083&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85101478083&partnerID=8YFLogxK
U2 - 10.1109/TWC.2020.3031503
DO - 10.1109/TWC.2020.3031503
M3 - Review article
AN - SCOPUS:85101478083
VL - 20
SP - 1188
EP - 1200
JO - IEEE Transactions on Wireless Communications
JF - IEEE Transactions on Wireless Communications
SN - 1536-1276
IS - 2
M1 - 9237168
ER -