Submit

Login Login

Register Register

Home / Journals / CMC / Online First / doi:10.32604/cmc.2025.063734
Journal Menu
Special Issues
Table of Content

All issues

Online First

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

Open Access

ARTICLE

Securing Internet of Things Devices with Federated Learning: A Privacy-Preserving Approach for Distributed Intrusion Detection

Sulaiman Al Amro*
Department of Computer Science, College of Computer, Qassim University, Buraydah, 51452, Saudi Arabia
* Corresponding Author: Sulaiman Al Amro. Email: email
(This article belongs to the Special Issue: Empowered Connected Futures of AI, IoT, and Cloud Computing in the Development of Cognitive Cities)

Computers, Materials & Continua https://doi.org/10.32604/cmc.2025.063734

Received 22 January 2025; Accepted 28 March 2025; Published online 21 April 2025

Abstract

The rapid proliferation of Internet of Things (IoT) devices has heightened security concerns, making intrusion detection a pivotal challenge in safeguarding these networks. Traditional centralized Intrusion Detection Systems (IDS) often fail to meet the privacy requirements and scalability demands of large-scale IoT ecosystems. To address these challenges, we propose an innovative privacy-preserving approach leveraging Federated Learning (FL) for distributed intrusion detection. Our model eliminates the need for aggregating sensitive data on a central server by training locally on IoT devices and sharing only encrypted model updates, ensuring enhanced privacy and scalability without compromising detection accuracy. Key innovations of this research include the integration of advanced deep learning techniques for real-time threat detection with minimal latency and a novel model to fortify the system’s resilience against diverse cyber-attacks such as Distributed Denial of Service (DDoS) and malware injections. Our evaluation on three benchmark IoT datasets demonstrates significant improvements: achieving 92.78% accuracy on NSL-KDD, 91.47% on BoT-IoT, and 92.05% on UNSW-NB15. The precision, recall, and F1-scores for all datasets consistently exceed 91%. Furthermore, the communication overhead was reduced to 85 MB for NSL-KDD, 105 MB for BoT-IoT, and 95 MB for UNSW-NB15—substantially lower than traditional centralized IDS approaches. This study contributes to the domain by presenting a scalable, secure, and privacy-preserving solution tailored to the unique characteristics of IoT environments. The proposed framework is adaptable to dynamic and heterogeneous settings, with potential applications extending to other privacy-sensitive domains. Future work will focus on enhancing the system’s efficiency and addressing emerging challenges such as model poisoning attacks in federated environments.

Keywords

Federated learning; internet of things; intrusion detection; privacy-preserving; distributed security

  • 365

    View

  • 196

    Download

  • 0

    Like

Related articles
Copyright© 2025 Tech Science Press
© 1997-2025 TSP (Henderson, USA) unless otherwise stated

Share Link