Hierarchical Over-the-Air Federated Learning with Differential Privacy

Zixi Wang; Arick Grootveld; M. Cenk Gursoy

doi:10.1145/3586209.3591400

Hierarchical Over-the-Air Federated Learning with Differential Privacy

Zixi Wang, Arick Grootveld, M. Cenk Gursoy

Department of Electrical Engineering & Computer Science

Research output: Chapter in Book/Entry/Poem › Conference contribution

Abstract

Federated learning (FL) is a burgeoning field that examines the cooperative interaction of machine learning (ML) models with users, enabling the training of a global model while each user retains its data locally. With differential privacy (DP), FL also becomes an enabler for training ML models in a more private manner. While there has been a growing body of work exploring various aspects of FL, most studies, especially in the context of hierarchical federated learning (HFL), treat different levels of the hierarchy as a composition of two DP mechanisms. In this paper, we introduce a DP based privacy preserving method with hierarchical over-the-air FL and address both communication and privacy aspects in an end-to-end fashion.

Original language	English (US)
Title of host publication	WiseML 2023 - Proceedings of the 2023 ACM Workshop on Wireless Security and Machine Learning
Publisher	Association for Computing Machinery, Inc
Pages	51-56
Number of pages	6
ISBN (Electronic)	9798400701337
DOIs	https://doi.org/10.1145/3586209.3591400
State	Published - Jun 1 2023
Event	5th ACM Workshop on Wireless Security and Machine Learning, WiseML 2023 - Guildford, United Kingdom Duration: Jun 1 2023 → …

Publication series

Name	WiseML 2023 - Proceedings of the 2023 ACM Workshop on Wireless Security and Machine Learning

Conference

Conference	5th ACM Workshop on Wireless Security and Machine Learning, WiseML 2023
Country/Territory	United Kingdom
City	Guildford
Period	6/1/23 → …

Keywords

differential privacy
federated learning
hierarchy

ASJC Scopus subject areas

Artificial Intelligence
Computer Networks and Communications
Software

Access to Document

10.1145/3586209.3591400

Cite this

Wang, Z., Grootveld, A., & Gursoy, M. C. (2023). Hierarchical Over-the-Air Federated Learning with Differential Privacy. In WiseML 2023 - Proceedings of the 2023 ACM Workshop on Wireless Security and Machine Learning (pp. 51-56). (WiseML 2023 - Proceedings of the 2023 ACM Workshop on Wireless Security and Machine Learning). Association for Computing Machinery, Inc. https://doi.org/10.1145/3586209.3591400

Hierarchical Over-the-Air Federated Learning with Differential Privacy. / Wang, Zixi; Grootveld, Arick; Gursoy, M. Cenk.
WiseML 2023 - Proceedings of the 2023 ACM Workshop on Wireless Security and Machine Learning. Association for Computing Machinery, Inc, 2023. p. 51-56 (WiseML 2023 - Proceedings of the 2023 ACM Workshop on Wireless Security and Machine Learning).

Research output: Chapter in Book/Entry/Poem › Conference contribution

Wang, Z, Grootveld, A & Gursoy, MC 2023, Hierarchical Over-the-Air Federated Learning with Differential Privacy. in WiseML 2023 - Proceedings of the 2023 ACM Workshop on Wireless Security and Machine Learning. WiseML 2023 - Proceedings of the 2023 ACM Workshop on Wireless Security and Machine Learning, Association for Computing Machinery, Inc, pp. 51-56, 5th ACM Workshop on Wireless Security and Machine Learning, WiseML 2023, Guildford, United Kingdom, 6/1/23. https://doi.org/10.1145/3586209.3591400

Wang Z, Grootveld A, Gursoy MC. Hierarchical Over-the-Air Federated Learning with Differential Privacy. In WiseML 2023 - Proceedings of the 2023 ACM Workshop on Wireless Security and Machine Learning. Association for Computing Machinery, Inc. 2023. p. 51-56. (WiseML 2023 - Proceedings of the 2023 ACM Workshop on Wireless Security and Machine Learning). doi: 10.1145/3586209.3591400

Wang, Zixi ; Grootveld, Arick ; Gursoy, M. Cenk. / Hierarchical Over-the-Air Federated Learning with Differential Privacy. WiseML 2023 - Proceedings of the 2023 ACM Workshop on Wireless Security and Machine Learning. Association for Computing Machinery, Inc, 2023. pp. 51-56 (WiseML 2023 - Proceedings of the 2023 ACM Workshop on Wireless Security and Machine Learning).

@inproceedings{26aff98e31b0438b90bb6b57e2799215,

title = "Hierarchical Over-the-Air Federated Learning with Differential Privacy",

abstract = "Federated learning (FL) is a burgeoning field that examines the cooperative interaction of machine learning (ML) models with users, enabling the training of a global model while each user retains its data locally. With differential privacy (DP), FL also becomes an enabler for training ML models in a more private manner. While there has been a growing body of work exploring various aspects of FL, most studies, especially in the context of hierarchical federated learning (HFL), treat different levels of the hierarchy as a composition of two DP mechanisms. In this paper, we introduce a DP based privacy preserving method with hierarchical over-the-air FL and address both communication and privacy aspects in an end-to-end fashion.",

keywords = "differential privacy, federated learning, hierarchy",

author = "Zixi Wang and Arick Grootveld and Gursoy, {M. Cenk}",

note = "Publisher Copyright: {\textcopyright} 2023 ACM.; 5th ACM Workshop on Wireless Security and Machine Learning, WiseML 2023 ; Conference date: 01-06-2023",

year = "2023",

month = jun,

day = "1",

doi = "10.1145/3586209.3591400",

language = "English (US)",

series = "WiseML 2023 - Proceedings of the 2023 ACM Workshop on Wireless Security and Machine Learning",

publisher = "Association for Computing Machinery, Inc",

pages = "51--56",

booktitle = "WiseML 2023 - Proceedings of the 2023 ACM Workshop on Wireless Security and Machine Learning",

}

TY - GEN

T1 - Hierarchical Over-the-Air Federated Learning with Differential Privacy

AU - Wang, Zixi

AU - Grootveld, Arick

AU - Gursoy, M. Cenk

PY - 2023/6/1

Y1 - 2023/6/1

N2 - Federated learning (FL) is a burgeoning field that examines the cooperative interaction of machine learning (ML) models with users, enabling the training of a global model while each user retains its data locally. With differential privacy (DP), FL also becomes an enabler for training ML models in a more private manner. While there has been a growing body of work exploring various aspects of FL, most studies, especially in the context of hierarchical federated learning (HFL), treat different levels of the hierarchy as a composition of two DP mechanisms. In this paper, we introduce a DP based privacy preserving method with hierarchical over-the-air FL and address both communication and privacy aspects in an end-to-end fashion.

AB - Federated learning (FL) is a burgeoning field that examines the cooperative interaction of machine learning (ML) models with users, enabling the training of a global model while each user retains its data locally. With differential privacy (DP), FL also becomes an enabler for training ML models in a more private manner. While there has been a growing body of work exploring various aspects of FL, most studies, especially in the context of hierarchical federated learning (HFL), treat different levels of the hierarchy as a composition of two DP mechanisms. In this paper, we introduce a DP based privacy preserving method with hierarchical over-the-air FL and address both communication and privacy aspects in an end-to-end fashion.

KW - differential privacy

KW - federated learning

KW - hierarchy

UR - http://www.scopus.com/inward/record.url?scp=85165946076&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85165946076&partnerID=8YFLogxK

U2 - 10.1145/3586209.3591400

DO - 10.1145/3586209.3591400

M3 - Conference contribution

AN - SCOPUS:85165946076

T3 - WiseML 2023 - Proceedings of the 2023 ACM Workshop on Wireless Security and Machine Learning

SP - 51

EP - 56

BT - WiseML 2023 - Proceedings of the 2023 ACM Workshop on Wireless Security and Machine Learning

PB - Association for Computing Machinery, Inc

T2 - 5th ACM Workshop on Wireless Security and Machine Learning, WiseML 2023

Y2 - 1 June 2023

ER -

Hierarchical Over-the-Air Federated Learning with Differential Privacy

Abstract

Publication series

Conference

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this