Blockchain-enhanced federated learning for IoT security and privacy using the GSR-C2N model

Abstract

Abstract The rapid proliferation of Internet of Things devices has intensified the demand for security, privacy-preserving, and scalable machine learning solutions. Federated Learning (FL) supports the decentralized training of models across distributed devices without transporting raw data, whereas blockchain provides a trusted, transparent mechanism for integrity and reaching consensus. The paper explains an FL framework that incorporates blockchain technology, based on the GSR-C2N model for identifying crypto-mining malware. It is demonstrated that the system’s feature extraction and optimization processes are optimized to address security problems in IoT by utilizing blockchain to verify model updates and build trust and privacy. The given structure has demonstrated superiority to the current practices. This model was 96.85% accurate and 97.51% specific on the crypto-mining malware data set using 10-fold cross-validation, making it applicable to smart healthcare and smart city applications with IoT-based systems. In addition, when combined with regulatory compliance, homomorphic encryption can strengthen data and privacy management, underscoring the model’s effectiveness in advanced IoT systems.