Privacy-Preserving Federated Learning in IoT Networks using Homomorphic Encryption and Differential Privacy

Fawad Khan; Syed Yaseen Shah; Syed Aziz Shah; Jawad Ahmad; Shahzaib Tahir; Adnan Zahid

doi:10.1109/ITC-Egypt66095.2025.11186608

Privacy-Preserving Federated Learning in IoT Networks using Homomorphic Encryption and Differential Privacy

Fawad Khan
, Syed Yaseen Shah
, Syed Aziz Shah
, Jawad Ahmad
, Shahzaib Tahir
, Adnan Zahid

School of Engineering & Physical Sciences

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Abstract

With the advancement of the Internet of Things (IoT) and big data analytics, organisations are focusing on Federated Learning (FL) for training models that incorporate data and device heterogeneity. Differential Privacy (DP) and Homomorphic Encryption (HE) are promising candidates for preserving privacy in FL, where DP focuses on adding noise and HE is centred on working with encrypted data. In this paper, we compare, analyse, and present the communication, computation, and accuracy aspects of both techniques for secure Federated Learning (FL) in IoT settings, under an honest-but-curious threat model. This paper also presents the usability of both techniques based on several metrics.

Original language	English
Title of host publication	2025 International Telecommunications Conference
Publisher	IEEE
Pages	131-136
Number of pages	6
ISBN (Electronic)	9781665458009
DOIs	https://doi.org/10.1109/ITC-Egypt66095.2025.11186608
Publication status	Published - 13 Oct 2025
Event	2025 International Telecommunications Conference - Cairo, Egypt Duration: 28 Jul 2025 → 31 Jul 2025

Conference

Conference	2025 International Telecommunications Conference
Abbreviated title	ITC-Egypt 2025
Country/Territory	Egypt
City	Cairo
Period	28/07/25 → 31/07/25

Keywords

cloud computing
logistic regression
model training
Paillier encryption
privacy preservation

ASJC Scopus subject areas

Artificial Intelligence
Computer Networks and Communications
Signal Processing
Safety, Risk, Reliability and Quality
Control and Optimization

Access to Document

10.1109/ITC-Egypt66095.2025.11186608

Cite this

@inproceedings{2cbaf90378f84bcc8cf2d3d274d49cc1,

title = "Privacy-Preserving Federated Learning in IoT Networks using Homomorphic Encryption and Differential Privacy",

abstract = "With the advancement of the Internet of Things (IoT) and big data analytics, organisations are focusing on Federated Learning (FL) for training models that incorporate data and device heterogeneity. Differential Privacy (DP) and Homomorphic Encryption (HE) are promising candidates for preserving privacy in FL, where DP focuses on adding noise and HE is centred on working with encrypted data. In this paper, we compare, analyse, and present the communication, computation, and accuracy aspects of both techniques for secure Federated Learning (FL) in IoT settings, under an honest-but-curious threat model. This paper also presents the usability of both techniques based on several metrics.",

keywords = "cloud computing, logistic regression, model training, Paillier encryption, privacy preservation",

author = "Fawad Khan and Shah, \{Syed Yaseen\} and Shah, \{Syed Aziz\} and Jawad Ahmad and Shahzaib Tahir and Adnan Zahid",

note = "Publisher Copyright: {\textcopyright} 2025 IEEE.; 2025 International Telecommunications Conference, ITC-Egypt 2025 ; Conference date: 28-07-2025 Through 31-07-2025",

year = "2025",

month = oct,

day = "13",

doi = "10.1109/ITC-Egypt66095.2025.11186608",

language = "English",

pages = "131--136",

booktitle = "2025 International Telecommunications Conference",

publisher = "IEEE",

address = "United States",

}

Khan, F, Shah, SY, Shah, SA, Ahmad, J, Tahir, S & Zahid, A 2025, Privacy-Preserving Federated Learning in IoT Networks using Homomorphic Encryption and Differential Privacy. in 2025 International Telecommunications Conference. IEEE, pp. 131-136, 2025 International Telecommunications Conference, Cairo, Egypt, 28/07/25. https://doi.org/10.1109/ITC-Egypt66095.2025.11186608

TY - GEN

T1 - Privacy-Preserving Federated Learning in IoT Networks using Homomorphic Encryption and Differential Privacy

AU - Khan, Fawad

AU - Shah, Syed Yaseen

AU - Shah, Syed Aziz

AU - Ahmad, Jawad

AU - Tahir, Shahzaib

AU - Zahid, Adnan

PY - 2025/10/13

Y1 - 2025/10/13

N2 - With the advancement of the Internet of Things (IoT) and big data analytics, organisations are focusing on Federated Learning (FL) for training models that incorporate data and device heterogeneity. Differential Privacy (DP) and Homomorphic Encryption (HE) are promising candidates for preserving privacy in FL, where DP focuses on adding noise and HE is centred on working with encrypted data. In this paper, we compare, analyse, and present the communication, computation, and accuracy aspects of both techniques for secure Federated Learning (FL) in IoT settings, under an honest-but-curious threat model. This paper also presents the usability of both techniques based on several metrics.

AB - With the advancement of the Internet of Things (IoT) and big data analytics, organisations are focusing on Federated Learning (FL) for training models that incorporate data and device heterogeneity. Differential Privacy (DP) and Homomorphic Encryption (HE) are promising candidates for preserving privacy in FL, where DP focuses on adding noise and HE is centred on working with encrypted data. In this paper, we compare, analyse, and present the communication, computation, and accuracy aspects of both techniques for secure Federated Learning (FL) in IoT settings, under an honest-but-curious threat model. This paper also presents the usability of both techniques based on several metrics.

KW - cloud computing

KW - logistic regression

KW - model training

KW - Paillier encryption

KW - privacy preservation

UR - https://www.scopus.com/pages/publications/105020912539

U2 - 10.1109/ITC-Egypt66095.2025.11186608

DO - 10.1109/ITC-Egypt66095.2025.11186608

M3 - Conference contribution

AN - SCOPUS:105020912539

SP - 131

EP - 136

BT - 2025 International Telecommunications Conference

PB - IEEE

T2 - 2025 International Telecommunications Conference

Y2 - 28 July 2025 through 31 July 2025

ER -

Privacy-Preserving Federated Learning in IoT Networks using Homomorphic Encryption and Differential Privacy

Abstract

Conference

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this