A Privacy-Preserving Consumer-Centric IoMT Framework Using TEE-Enabled Federated Learning, CP-ABE, and Blockchain

Abstract

The rapid expansion of the Internet of Medical Things (IoMT) has transformed healthcare delivery by enabling real-time monitoring, advanced diagnostics, and efficient data sharing. However, current systems in large-scale, decentralized environments face significant challenges in privacy, security, trust, and interoperability. This paper presents a Federated Learning framework for a consumer-centric IoMT system that provides secure, resilient data exchange against AI-enabled attacks through privacy-preserving learning, decentralized authentication, and hardware-based trust. It integrates trusted execution environments (TEEs), blockchain-based trust management, and ciphertext-policy attribute-based encryption (CP-ABE) to enable secure collaboration without exposing raw data. We provide a security proof in the Real-or-Random (RoR) model and conduct a comprehensive performance analysis covering computation, communication, storage, smart contract processing, and throughput. Experimental results show a 40% reduction in authentication latency, a 27% decrease in computational overhead, a 98% drop in energy use, and a 35% increase in throughput compared to existing schemes. These findings demonstrate that the proposed framework offers high scalability robust security, and privacy protection, making it ideal for the next-generation healthcare ecosystem.