Differentially Private Federated Learning Against Multi-Antenna Passive Eavesdroppers 


Vol. 50,  No. 9, pp. 1444-1446, Sep.  2025
10.7840/kics.2025.50.9.1444


PDF Full-Text
  Abstract

In this work, we study differentially private wireless federated learning (FL) in the presence of a multi-antenna passive eavesdropper. Specifically, we consider an artificial noise (AN) scheme in which a multi-antenna FL user utilizes beamforming to inject AN signal, aiming to minimize inference accuracy degradation caused by ensuring differential privacy (DP). Unlike existing studies that require wireless channel information between the FL user and the eavesdropper to guarantee DP, this work mathe-matically shows that DP can be ensured without the channel information and propose an optimal AN scheme to achieve DP.

  Statistics
Cumulative Counts from November, 2022
Multiple requests among the same browser session are counted as one view. If you mouse over a chart, the values of data points will be shown.


  Related Articles
  Cite this article

[IEEE Style]

J. Park and S. Yun, "Differentially Private Federated Learning Against Multi-Antenna Passive Eavesdroppers," The Journal of Korean Institute of Communications and Information Sciences, vol. 50, no. 9, pp. 1444-1446, 2025. DOI: 10.7840/kics.2025.50.9.1444.

[ACM Style]

Junguk Park and Sangseok Yun. 2025. Differentially Private Federated Learning Against Multi-Antenna Passive Eavesdroppers. The Journal of Korean Institute of Communications and Information Sciences, 50, 9, (2025), 1444-1446. DOI: 10.7840/kics.2025.50.9.1444.

[KICS Style]

Junguk Park and Sangseok Yun, "Differentially Private Federated Learning Against Multi-Antenna Passive Eavesdroppers," The Journal of Korean Institute of Communications and Information Sciences, vol. 50, no. 9, pp. 1444-1446, 9. 2025. (https://doi.org/10.7840/kics.2025.50.9.1444)
Vol. 50, No. 9 Index

Search
(IN TITLE, AUTHOR, ABSTRACT,KEYWORDS)

Advanced Search ▼

POPULAR KEYWORDS
(TOP 10 KEYWORDS)

Recent Publications
(LAST 3 YEARS)






pISSN : 1226 - 4717
eISSN : 2287 - 3880