Enabling Efficient Multi-keyword Search Over Fine-Grained Authorized Healthcare Blockchain System

Abstract

As a new emerging technology, blockchain is attracting the attention from academic and industry and has been widely exploited to build the large-scale data sharing and management systems, such as healthcare database or bank distributed database system. The health records contain a lot of sensitive information, so putting these health records into blockchain can solve the security and privacy issues while uploading them to an untrustworthy network. In a typical health record management system, there are escalating demands for users including the patients and the doctors to execute multi-keyword search over the huge scale of healthcare records. In the meantime, they can authorize some part of their personal treatments to others according to personalized needs of the patients. In literatures, there is not an existing blockchain solution can satisfy these two requirements at the same time. These issues become prominent since it’s more inconvenient to adjust a blockchain-based system to support efficient multi-keyword search and fine-grained authorization comparing to traditional RDBMS. To overcome the two challenges, we propose a novel multi-keyword searching scheme by establishing a set of Bloom Filters within the health record blockchain system to accelerate the searching process on service provider (SP). Moreover, we reduce the overhead of key derivation by proposing a Healthcare Data Key Derivation Tree (HDKDT) stored locally on the user’s side. Putting our proposed scheme on the medical blockchain can speed up the multi-keyword search [, ] processes and reduce the key storage space to certain extent. At the end of this article, we formally prove the security of the proposed scheme and implement a prototype system to evaluate its performance. The experimental results validate our proposed scheme in this paper is a secure and efficient approach for the health record management scenario.