Efficient Policy-Hiding and Large Universe Attribute-Based Encryption with Public Traceability for Internet of Medical Things

Peng Zeng,Zhiting Zhang,Rongxing Lu,Kim-Kwang Raymond Choo
DOI: https://doi.org/10.1109/jiot.2021.3051362
IF: 10.6
2021-01-01
IEEE Internet of Things Journal
Abstract:Modern day medical systems are closely integrated and interconnected with other systems, such as those comprising Internet-of-Medical Things (IoMT) devices that facilitate remote healthcare services, say during pandemics (e.g., COVID-19). Attribute-based encryption (ABE) is a promising cryptographic primitive to support fine-grained access control in the ciphertext environment; in other words, ABE can potentially be used to ensure data confidentiality and user privacy in the IoMT ecosystem. In this article, we propose an efficient partially-policy-hidden and large universe ABE scheme with public traceability to construct a practical IoMT system (hereafter referred to as PTIoMT). The system is designed to achieve the following features: 1) the access policy is partially hidden: only nonsensitive attribute labels/names are displayed, while sensitive attribute values are hidden in the encrypted electronic health records (EHRs); 2) the number of the attributes is independent of the public parameters and, thus, can be arbitrarily large; 3) any user who discloses the decryption key can be efficiently tracked; and 4) fewer bilinear pairing operations are required during the decryption process. The security analysis and performance evaluation demonstrate the security and efficiency of PTIoMT.
What problem does this paper attempt to address?