EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

TinyFed : Lightweight Federated Learning for Constrained IoT Devices

Claudio Jr. N. da Silva, Maycon L. M. Peixoto, Gustavo B. Figueiredo, Cassio V. S. Prazeres

Internet Technology Letters · 2025-07 · 8 (4)

DOI: 10.1002/itl2.70061 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Internet Technology Letters, online erschienen am 22.06.2025. Verfasst von Claudio Jr. N. da Silva und 3 weiteren Autor:innen. Die Arbeit wurde bisher 7-mal zitiert (Crossref) und nennt 17 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

TinyML enables machine learning inference on microcontrollers with limited resources. Extending this to a collaborative setting led to Tiny Federated Learning (TinyFL). This article presents TinyFed, a lightweight framework that supports the full federated learning cycle—from local training to model aggregation and redistribution. TinyFed was validated on ESP32 devices using a neural network with four inputs, three hidden layers, and two outputs to detect temperature, humidity, luminosity, and voltage anomalies. Local training achieved accuracies of up to 99.47%.

Zitieren

APA

Silva, C. J. N. D., Peixoto, M. L. M., Figueiredo, G. B., & Prazeres, C. V. S. (2025). TinyFed : Lightweight Federated Learning for Constrained IoT Devices. Internet Technology Letters, 8(4). https://doi.org/10.1002/itl2.70061

Vancouver

Silva CJND, Peixoto MLM, Figueiredo GB, Prazeres CVS. TinyFed : Lightweight Federated Learning for Constrained IoT Devices. Internet Technology Letters. 2025;8(4). doi:10.1002/itl2.70061

BibTeX
@article{Silva2025,
  author = {Silva, Claudio Jr. N. da and Peixoto, Maycon L. M. and Figueiredo, Gustavo B. and Prazeres, Cassio V. S.},
  title = {TinyFed : Lightweight Federated Learning for Constrained IoT Devices},
  journal = {Internet Technology Letters},
  year = {2025},
  volume = {8},
  number = {4},
  doi = {10.1002/itl2.70061}
}
Kostenlos starten →