EUVIMEDEUVIMED entdecken
Crossref

FedVQCS: Federated Learning via Vector Quantized Compressed Sensing

Yongjeong Oh, Yo-Seb Jeon, Mingzhe Chen, Walid Saad

IEEE Transactions on Wireless Communications · 2024-03 · 23 (3) · 1755-1770

DOI: 10.1109/twc.2023.3291877 ↗

Kurz gefasst

Fachartikel in der Zeitschrift IEEE Transactions on Wireless Communications, 2024-03. Verfasst von Yongjeong Oh und 3 weiteren Autor:innen. Die Arbeit wurde bisher 35-mal zitiert (Crossref) und nennt 45 Literaturangaben. Gefördert von: National Research Foundation of Korea (NRF) Grant, Korea Government through MSIT, Institute of Information and Communications Technology Planning and Evaluation (IITP) Grant.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

In this paper, a new communication-efficient federated learning (FL) framework is proposed, inspired by vector quantized compressed sensing. The basic strategy of the proposed framework is to compress the local model update at each device by applying dimensionality reduction followed by vector quantization. Subsequently, the global model update is reconstructed at a parameter server by applying a sparse signal recovery algorithm to the aggregation of the compressed local model updates. By harnessing the benefits of both dimensionality reduction and vector quantization, the proposed framework effectively reduces the communication overhead of local update transmissions. Both the design of the vector quantizer and the key parameters for the compression are optimized so as to minimize the reconstruction error of the global model update under the constraint of wireless link capacity. By considering the reconstruction error, the convergence rate of the proposed framework is also analyzed for a non-convex loss function. Simulation results on the MNIST and FEMNIST datasets demonstrate that the proposed framework can improve classification accuracy by more than 2.4% compared to state-of-the-art FL frameworks when the communication overhead of the local model update transmission is 0.1 bit per local model entry.

Förderung

  • National Research Foundation of Korea (NRF) Grant
  • Korea Government through MSIT
  • Institute of Information and Communications Technology Planning and Evaluation (IITP) Grant
  • Korea Government through MSIT [Development of Intelligent Wireless Access Technology, (50%)]
  • BK21 FOUR Program
  • U.S. National Science Foundation

Zitieren

APA

Oh, Y., Jeon, Y., Chen, M., & Saad, W. (2024). FedVQCS: Federated Learning via Vector Quantized Compressed Sensing. IEEE Transactions on Wireless Communications, 23(3), 1755-1770. https://doi.org/10.1109/twc.2023.3291877

Vancouver

Oh Y, Jeon Y, Chen M, Saad W. FedVQCS: Federated Learning via Vector Quantized Compressed Sensing. IEEE Transactions on Wireless Communications. 2024;23(3):1755-1770. doi:10.1109/twc.2023.3291877

BibTeX
@article{Oh2024,
  author = {Oh, Yongjeong and Jeon, Yo-Seb and Chen, Mingzhe and Saad, Walid},
  title = {FedVQCS: Federated Learning via Vector Quantized Compressed Sensing},
  journal = {IEEE Transactions on Wireless Communications},
  year = {2024},
  volume = {23},
  number = {3},
  pages = {1755--1770},
  doi = {10.1109/twc.2023.3291877}
}
Kostenlos starten →