A secure and scalable Federated Learning–based architecture with a lightweight authentication protocol to enable collaborative learning from distributed IoT data streams and a federated version of the Hoeffding tree algorithm for real-time learning is developed.
The proposed framework introduces several innovative features, such as federated learning with momentum-based optimization, adaptive differential privacy, trust verification via blockchain, and Byzantine-resilient aggregation, to enhance the security, scalability, and robustness of the system compared with traditional...
The experimental results show that the proposed AKA+ECC protocol is superior to the existing DAC+ECC, LAP+ECC, and SLA+ECC authentication schemes, with a computational cost of 8.3 ms and a communication overhead of 1044 bits, and is resistant to replay attacks, impersonation attacks, and man-in-the-middle attack.
Aruna Thota, Ch.Jayanth Babu, G. Rawat· 2026 6th International Confe...· 0 citations
The findings of this review show that proactive, cooperative, and self-healing security mechanisms are more effective than traditional reactive approaches in IoT environments and provide recommendations for developing scalable, energy-efficient, and privacy-preserving IoT security frameworks for future research.
A federated learning framework to collaboratively develop an anomaly detection model from multiple edge nodes, without sharing the raw sensor data, so that data sovereignty is preserved, and demonstrates that practical, deployable security for federated learning in IIoT is achievable without sacrificing accuracy or res...
Resource-constrained Internet of Things (IoT) devices and wireless sensor networks (WSNs) remain difficult to secure using conventional security mechanisms because their sensing, forwarding, and processing nodes operate with limited memory, computation, energy, bandwidth, and software-update capability. These limits ex...
O. Khashan· Discover Internet of Things· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.