A TEE-based multi-data-owner collaborative attribute revocation scheme for secure medical data sharing

With the growing demand for cross-institutional data sharing, achieving secure collaboration among multiple data owners while preserving privacy has become a critical challenge in data security research. Traditional attribute-based encryption (ABE)–based access control mechanisms suffer from high computational overhead, frequent ciphertext re-encryption, and heavy reliance on centralized authorities when handling attribute revocation and multi-party coordination, making them unsuitable for dynamic security requirements in highly sensitive domains such as healthcare. To address these limitations, this paper proposes a TEE-based controllable data-sharing mechanism for multiple data owners. Leveraging the hardware-level isolation and trusted computing capabilities of the Trusted Execution Environment (TEE), the proposed mechanism encapsulates key operations within the secure execution domain, thereby significantly improving the efficiency of attribute revocation and reducing computational and communication costs. Furthermore, a collaborative attribute revocation mechanism is designed, enabling multiple data owners to jointly participate in access authorization decisions and realize decentralized dynamic access control. Extensive experimental evaluations demonstrate that the proposed mechanism achieves strong data confidentiality and fine-grained controllability with high efficiency and scalability, showing its potential for secure and privacy-preserving data sharing in sensitive healthcare environments.