EA-AQF: energy aware adaptive quantization and freezing in federated learning
Cluster Computing · 2026-09 · 29 (12)
DOI: 10.1007/s10586-026-06503-0 ↗Kurz gefasst
Fachartikel in der Zeitschrift Cluster Computing, online erschienen am 25.08.2026. Verfasst von Farwa Ikram und 4 weiteren Autor:innen. Die Arbeit wurde bisher 1-mal zitiert (Crossref) und nennt 49 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Blekinge Institute of Technology.
Abstract (Originalsprache)
Federated Learning (FL) enables collaborative model training across decentralized edge devices while preserving data privacy. However, real-world deployments are severely hindered by high uplink communication overhead and heterogeneous on-device computational constraints. To address these joint bottlenecks, We present Energy-Aware Adaptive Quantization and Freezing (EA-AQF), a unified framework that co-optimizes communication and computation. EA-AQF introduces a tiered mapping policy that assigns stage-wise activation bitwidths based on a dynamic client resource indicator. To further reduce local training costs, we implement a deterministic, precision-conditioned stage-freezing policy that disables gradient updates for computationally expensive network stages during backpropagation. Additionally, weight quantization is applied to client updates to minimize the communication footprint. We evaluate EA-AQF across diverse architectures, including CNN-7, CNN-17, and ResNet-20 on CIFAR-10, CIFAR-100, and MNIST datasets, and LSTM and MLP on IMDB and HAR datasets, under varying statistical distributions. Experimental results demonstrate that EA-AQF achieves a consistent communication reduction of 72.9% to 74.8% and significant computation reduction up to 12.66 $$\times$$ , which is 1.6 $$\times$$ higher than prior state-of-the-art methods, with a marginal accuracy trade-off typically within 2-3% of full-precision baselines. Quantitatively, our sustainability analysis reveals that EA-AQF achieves up to 65% reduction in energy consumption compared to state-of-the-art, while strictly limiting carbon emissions to as low as 0.0122 $$\mathrm {kgCO_2e}$$ for resource-intensive tasks. Notably, in highly heterogeneous tasks, EA-AQF maintains robust convergence positioning it as a highly efficient solution for sustainable edge AI.
Förderung
- Blekinge Institute of Technology
Zitieren
APA
Ikram, F., Alawadi, S., Thakur, D., Guzzo, A., & Fortino, G. (2026). EA-AQF: energy aware adaptive quantization and freezing in federated learning. Cluster Computing, 29(12). https://doi.org/10.1007/s10586-026-06503-0
Vancouver
Ikram F, Alawadi S, Thakur D, Guzzo A, Fortino G. EA-AQF: energy aware adaptive quantization and freezing in federated learning. Cluster Computing. 2026;29(12). doi:10.1007/s10586-026-06503-0
BibTeX
@article{Ikram2026,
author = {Ikram, Farwa and Alawadi, Sadi and Thakur, Dipanwita and Guzzo, Antonella and Fortino, Giancarlo},
title = {EA-AQF: energy aware adaptive quantization and freezing in federated learning},
journal = {Cluster Computing},
year = {2026},
volume = {29},
number = {12},
doi = {10.1007/s10586-026-06503-0}
}
EUVIMED