Designing a Scalable and Secure IoT Framework Using Federated Learning and Blockchain for Edge-AI Devices

S. Kannadhasan, Pilli Lalitha Kumari, K. Suresh, Badepally Mallaiah, Abirami G., Syed Zahidur Rashid

2025

Abstract

The need for flexible, secure, and intelligent data processing at the edge has been propelled by the fast development of Internet of Things (IoT) ecosystems. Existing federated learning (FL) methods usually suffer from system heterogeneity, privacy threats, and excessive communication cost. Additionally, adopting blockchain technology within FL typically adds both latency and complexity which limits its practical applicability to resource-constrained environments. In this paper, we introduce Edge Secure-Fed Chain, a new lightweight and trust-aware federated learning framework that incorporates blockchain, designed to enable secure and decentralized coordination among edge-AI devices. In contrast to existing approaches, our architecture achieves low latency via protocol-optimizing consensus, enables dynamic smart contract driven ML workflows, and improves personalization through adaptive local training. We also propose a resilient multi-tiered aggregation system (against adversarial and non-IID data conditions), together with proactive defense components (network anomaly detection and client reputation scoring). Edge Secure- Fed Chain outperforms the existing systems by overcoming their limitations as illustrated in this paper, which exhibit to be more scalable, preserve privacy, and have real-time performance in edge oriented IoT applications. Extensive experimental assessments validate the framework's efficacy, security, and adaptability to various IoT applications.

Paper Citation

in Harvard Style

Kannadhasan S., Kumari P., Suresh K., Mallaiah B., G. A. and Rashid S. (2025). Designing a Scalable and Secure IoT Framework Using Federated Learning and Blockchain for Edge-AI Devices. In Proceedings of the 1st International Conference on Research and Development in Information, Communication, and Computing Technologies - Volume 1: ICRDICCT`25; ISBN 978-989-758-777-1, SciTePress, pages 598-607. DOI: 10.5220/0013870100004919

in Bibtex Style

@conference{icrdicct`2525,
author={S. Kannadhasan and Pilli Kumari and K. Suresh and Badepally Mallaiah and Abirami G. and Syed Rashid},
title={Designing a Scalable and Secure IoT Framework Using Federated Learning and Blockchain for Edge-AI Devices},
booktitle={Proceedings of the 1st International Conference on Research and Development in Information, Communication, and Computing Technologies - Volume 1: ICRDICCT`25},
year={2025},
pages={598-607},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0013870100004919},
isbn={978-989-758-777-1},
}

in EndNote Style

TY - CONF

JO - Proceedings of the 1st International Conference on Research and Development in Information, Communication, and Computing Technologies - Volume 1: ICRDICCT`25
TI - Designing a Scalable and Secure IoT Framework Using Federated Learning and Blockchain for Edge-AI Devices
SN - 978-989-758-777-1
AU - Kannadhasan S.
AU - Kumari P.
AU - Suresh K.
AU - Mallaiah B.
AU - G. A.
AU - Rashid S.
PY - 2025
SP - 598
EP - 607
DO - 10.5220/0013870100004919
PB - SciTePress