Privacy-Preserving Traceable Attribute-Based Keyword Search in Multi-Authority Medical Cloud

In cloud-based electronic medical record (EMR) systems, attribute-based encryption (ABE) has been utilized to protect the confidentiality of EMRs and provide keyword search over the encrypted EMRs. However, existing schemes are designed for a single attribute authority, and lack sufficient user privacy protection. In this article, we introduce TABKS, a privacy-preserving traceable attribute-based keyword search scheme in multi-authority medical cloud. First, we propose an anonymous EMR access control framework with multiple authorities, which provides user anonymity against the untrusted authorities. Second, we achieve traceable attribute-based Boolean keyword search, which enables the authorized user who satisfies the policy to conduct Boolean keyword search over the encrypted EMRs. In this process, TABKS improves the efficiency of legitimate users by partially decrypting the matched results, and also achieves efficient traitor trace by revealing the user identity from the trapdoor. Finally, we prove the security of TABKS against chosen plaintext attack and chosen keyword attack, and conduct extensive experiments with two real-world datasets to show the feasibility of TABKS.