EcoLearn: Optimizing the Carbon Footprint of Federated Learning
Not on the map. Why?
Abstract (arXiv)
Federated Learning (FL) distributes machine learning (ML) training across edge devices to reduce data transfer overhead and protect data privacy. Since FL model training may span hundreds of devices and is thus resource- and energy-intensive, it has a significant carbon footprint. Importantly, since energy's carbon-intensity differs substantially (by up to 60×) across locations, training on the same device using the same amount of energy, but at different locations, can incur widely different carbon emissions. While prior work has focused on improving FL's resource- and energy-efficiency by optimizing time-to-accuracy, it implicitly assumes all energy has the same carbon intensity and thus does not optimize carbon efficiency, i.e., work done per unit of carbon emitted. To address the problem, we design EcoLearn, which minimizes FL's carbon footprint without significantly affecting model accuracy or training time. EcoLearn achieves a favorable tradeoff by integrating carbon awareness into multiple aspects of FL training, including i) selecting clients with high data utility and low carbon, ii) provisioning more clients during the initial training rounds, and iii) mitigating stragglers by dynamically adjusting client over-provisioning based on carbon. We implement EcoLearn and its carbon-aware FL training policies in the Flower framework and show that it reduces the carbon footprint of training (by up to 10.8×) while maintaining model accuracy and training time (within ~1%) compared to state-of-the-art approaches.
Key facts. Cited by 1 paper in the map (data through 14 Oct 2026). Builds on 35 of its 62 references that are in the map. First posted on arXiv 27 Oct 2023.
Builds on
35 of 62 references.
References · 35
- 2015Deep Residual Learning for Image Recognitioncited by 52,096
- 2016Communication-Efficient Learning of Deep Networks from Decentralized Datacited by 7,669
- 2018Federated Optimization in Heterogeneous Networkscited by 2,987
- 2019Towards Federated Learning at Scale: System Designcited by 832
- 2018LEAF: A Benchmark for Federated Settingscited by 792
- 2018Speech Commands: A Dataset for Limited-Vocabulary Speech Recognitioncited by 752
- 2017Ray: A Distributed Framework for Emerging AI Applicationscited by 527
- 2019Local Model Poisoning Attacks to Byzantine-Robust Federated Learningcited by 493
- 2017Differentially Private Federated Learning: A Client Level Perspectivecited by 459
- 2020Flower: A Friendly Federated Learning Research Frameworkcited by 455
- 2018Adaptive Federated Learning in Resource Constrained Edge Computing Systemscited by 409
- 2019Fair Resource Allocation in Federated Learningcited by 394
23 more
- 2018Not All Samples Are Created Equal: Deep Learning with Importance Samplingcited by 328
- 2017Don't Decay the Learning Rate, Increase the Batch Sizecited by 320
- 2019The Non-IID Data Quagmire of Decentralized Machine Learningcited by 305
- 2021Sustainable AI: Environmental Implications, Challenges and Opportunitiescited by 242
- 2018Applied Federated Learning: Improving Google Keyboard Query Suggestionscited by 204
- 2022Will we run out of data? Limits of LLM scaling based on human-generated datacited by 195
- 2018Expanding the Reach of Federated Learning by Reducing Client Resource Requirementscited by 163
- 2020Optimal Client Sampling for Federated Learningcited by 124
- 2018LAG: Lazily Aggregated Gradient for Communication-Efficient Distributed Learningcited by 96
- 2019Federated Learning over Wireless Networks: Convergence Analysis and Resource Allocationcited by 94
- 2021Papaya: Practical, Private, and Scalable Federated Learningcited by 90
- 2018On the Relation Between the Sharpest Directions of DNN Loss and the SGD Step Lengthcited by 88
- 2022Client Selection in Federated Learning: Principles, Challenges, and Opportunitiescited by 62
- 2021Keyword Transformer: A Self-Attention Model for Keyword Spottingcited by 60
- 2019Time Matters in Regularizing Deep Networks: Weight Decay and Data Augmentation Affect Early Learning Dynamics, Matter Little Near Convergencecited by 54
- 2020Can Federated Learning Save The Planet?cited by 52
- 2022FedBalancer: Data and Pace Control for Efficient Federated Learning on Heterogeneous Clientscited by 46
- 2021Resource-Efficient Federated Learningcited by 26
- 2020Accordion: Adaptive Gradient Communication via Critical Learning Regime Identificationcited by 19
- 2023FedZero: Leveraging Renewable Excess Energy in Federated Learningcited by 11
- 2022A Snapshot of the Frontiers of Client Selection in Federated Learningcited by 9
- 2021Critical Learning Periods in Federated Learningcited by 8
- 2020Efficient Replication for Straggler Mitigation in Distributed Computingcited by 1
Built on by
Citing papers per year: 2025: 1.
1 paper.