TY - JOUR
T1 - Scheduling Policies for Federated Learning in Wireless Networks
AU - Yang, Howard H.
AU - Liu, Zuozhu
AU - Quek, Tony Q.S.
AU - Poor, H. Vincent
N1 - Funding Information:
This work was supported in part by the SUTD-ZJU Research Collaboration under Grant SUTD-ZJU/RES/01/2016, in part by the SUTD-ZJU Research Collaboration under Grant SUTD-ZJU/RES/05/2016, and in part by the SUTD Growth Plan Grant for AI. It was also supported in part by the U.S. National Science Foundation under Grants CCF-093970 and CCF-1513915. The associate editor coordinating the review of this article and approving it for publication was C. Fischione.
Funding Information:
Manuscript received April 7, 2019; revised August 17, 2019; accepted September 23, 2019. Date of publication September 27, 2019; date of current version January 15, 2020. This work was supported in part by the SUTD-ZJU Research Collaboration under Grant SUTD-ZJU/RES/01/2016, in part by the SUTD-ZJU Research Collaboration under Grant SUTD-ZJU/RES/05/2016, and in part by the SUTD Growth Plan Grant for AI. It was also supported in part by the U.S. National Science Foundation under Grants CCF-093970 and CCF-1513915. The associate editor coordinating the review of this article and approving it for publication was C. Fischione. (Corresponding author: Zuozhu Liu.) H. H. Yang and T. Q. S. Quek are with the Information System Technology and Design Pillar, Singapore University of Technology and Design, Singapore 487372 (e-mail: howard_yang@sutd.edu.sg; tonyquek@sutd.edu.sg).
Publisher Copyright:
© 2019 IEEE.
PY - 2020/1
Y1 - 2020/1
N2 - Motivated by the increasing computational capacity of wireless user equipments (UEs), e.g., smart phones, tablets, or vehicles, as well as the increasing concerns about sharing private data, a new machine learning model has emerged, namely federated learning (FL), that allows a decoupling of data acquisition and computation at the central unit. Unlike centralized learning taking place in a data center, FL usually operates in a wireless edge network where the communication medium is resource-constrained and unreliable. Due to limited bandwidth, only a portion of UEs can be scheduled for updates at each iteration. Due to the shared nature of the wireless medium, transmissions are subjected to interference and are not guaranteed. The performance of FL system in such a setting is not well understood. In this paper, an analytical model is developed to characterize the performance of FL in wireless networks. Particularly, tractable expressions are derived for the convergence rate of FL in a wireless setting, accounting for effects from both scheduling schemes and inter-cell interference. Using the developed analysis, the effectiveness of three different scheduling policies, i.e., random scheduling (RS), round robin (RR), and proportional fair (PF), are compared in terms of FL convergence rate. It is shown that running FL with PF outperforms RS and RR if the network is operating under a high signal-to-interference-plus-noise ratio (SINR) threshold, while RR is more preferable when the SINR threshold is low. Moreover, the FL convergence rate decreases rapidly as the SINR threshold increases, thus confirming the importance of compression and quantization of the update parameters. The analysis also reveals a trade-off between the number of scheduled UEs and subchannel bandwidth under a fixed amount of available spectrum.
AB - Motivated by the increasing computational capacity of wireless user equipments (UEs), e.g., smart phones, tablets, or vehicles, as well as the increasing concerns about sharing private data, a new machine learning model has emerged, namely federated learning (FL), that allows a decoupling of data acquisition and computation at the central unit. Unlike centralized learning taking place in a data center, FL usually operates in a wireless edge network where the communication medium is resource-constrained and unreliable. Due to limited bandwidth, only a portion of UEs can be scheduled for updates at each iteration. Due to the shared nature of the wireless medium, transmissions are subjected to interference and are not guaranteed. The performance of FL system in such a setting is not well understood. In this paper, an analytical model is developed to characterize the performance of FL in wireless networks. Particularly, tractable expressions are derived for the convergence rate of FL in a wireless setting, accounting for effects from both scheduling schemes and inter-cell interference. Using the developed analysis, the effectiveness of three different scheduling policies, i.e., random scheduling (RS), round robin (RR), and proportional fair (PF), are compared in terms of FL convergence rate. It is shown that running FL with PF outperforms RS and RR if the network is operating under a high signal-to-interference-plus-noise ratio (SINR) threshold, while RR is more preferable when the SINR threshold is low. Moreover, the FL convergence rate decreases rapidly as the SINR threshold increases, thus confirming the importance of compression and quantization of the update parameters. The analysis also reveals a trade-off between the number of scheduled UEs and subchannel bandwidth under a fixed amount of available spectrum.
KW - Federated learning
KW - convergence analysis
KW - parallel and distributed algorithms
KW - scheduling policies
KW - stochastic geometry
UR - http://www.scopus.com/inward/record.url?scp=85078238110&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85078238110&partnerID=8YFLogxK
U2 - 10.1109/TCOMM.2019.2944169
DO - 10.1109/TCOMM.2019.2944169
M3 - Article
AN - SCOPUS:85078238110
SN - 1558-0857
VL - 68
SP - 317
EP - 333
JO - IEEE Transactions on Communications
JF - IEEE Transactions on Communications
IS - 1
M1 - 8851249
ER -