Energy-Efficient Radio Resource Allocation for Federated Edge Learning

Zeng, Qunsong; Du, Youwei; Huang, Kaibin; Leung, Kin K.

doi:10.1109/iccworkshops49005.2020.9145118

Cited by 206 publications

(133 citation statements)

References 6 publications

Supporting

Mentioning

124

Contrasting

Order By: Relevance

“…To illustrate, in [1], when making a selection, Nishio et al concentrate on the evaluation of communication time, which accounts for a considerable portion of time for a training round. In another study [2], the authors consider more. They further take the energy consumption factor into consideration.…”

Section: Motivationsmentioning

confidence: 99%

“…For one thing, both of them assume a pre-known local training time to the scheduler, which may not be realistic in all circumstances. For another, indicated by Theorem 2 in [2], devices with higher performance are more favored by their proposed methods. Indeed, always selecting the "fast" devices somehow boost the training process.…”

Section: Motivationsmentioning

confidence: 99%

“…Another open problem of FL we would like to mention here is the client selection problem, originally proposed in [1] and followed by some related works (e.g. [2], [14]- [16]). Many of them see the problem from a communication perspective, focusing on building an efficient selection or bandwidth allocation scheme that helps shorten the communication length.…”

Section: Related Workmentioning

confidence: 99%

“…For clearness, one can check Table 1 for the inherent training setting of different classes of clients. For the context generation (in order to simulate the perround status of clients), we assume the allocated bandwidth of all clients is sampling from a uniform distribution between [2,4]MHz and the model size M is fixed to 20Mb. Likewise, the available computation capacity of all clients is also sampling from the same uniform distribution within [50%, 200%].…”

Section: Simulation Settingmentioning

confidence: 99%

See 3 more Smart Citations

An Efficiency-boosting Client Selection Scheme for Federated Learning with Fairness Guarantee

Huang

Lin

et al. 2020

IEEE Trans. Parallel Distrib. Syst.

135

View full text Add to dashboard Cite

The issue of potential privacy leakage during centralized AI's model training has drawn intensive concern from the public. A Parallel and Distributed Computing (or PDC) scheme, termed Federated Learning (FL), has emerged as a new paradigm to cope with the privacy issue by allowing clients to perform model training locally, without the necessity to upload their personal sensitive data. In FL, the number of clients could be sufficiently large, but the bandwidth available for model distribution and re-upload is quite limited, making it sensible to only involve part of the volunteers to participate in the training process. The client selection policy is critical to an FL process in terms of training efficiency, the final model's quality as well as fairness. In this paper, we will model the fairness guaranteed client selection as a Lyapunov optimization problem and then a C 2 MAB-based method is proposed for estimation of the model exchange time between each client and the server, based on which we design a fairness guaranteed algorithm termed RBCS-F for problem-solving. The regret of RBCS-F is strictly bounded by a finite constant, justifying its theoretical feasibility. Barring the theoretical results, more empirical data can be derived from our real training experiments on public datasets.

show abstract

Section: Motivationsmentioning

confidence: 99%

Section: Motivationsmentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

Section: Simulation Settingmentioning

confidence: 99%

See 2 more Smart Citations

An Efficiency-boosting Client Selection Scheme for Federated Learning with Fairness Guarantee

Huang

Lin

et al. 2020

IEEE Trans. Parallel Distrib. Syst.

135

View full text Add to dashboard Cite

show abstract

“…The authors demonstrate its effectiveness in terms of classification performance using three real-world datasets. Energy-efficient radio resource allocation for enabling FL in an edge scenario has been investigated in [35] by adapting the communication to devices' channel states and computation capacities so as to reduce their energy consumption while guaranteeing learning performance.…”

Section: Introductionmentioning

confidence: 99%