$π$QLB: A Privacy-preserving with Integrity-assuring Query Language for Blockchain (2212.14141v2)

Abstract: The increase in the adoption of blockchain technology in different application domains e.g., healthcare systems, supplychain management, has raised the demand for a data query mechanism on blockchain. Since current blockchain systems lack the support for querying data with embedded security and privacy guarantees, there exists inherent security and privacy concerns on those systems. In particular, existing systems require users to submit queries to blockchain operators (e.g., a node validator) in plaintext. This directly jeopardizes users' privacy as the submitted queries may contain sensitive information, e.g., location or gender preferences, that the users may not be comfortable sharing. On the other hand, currently, the only way for users to ensure integrity of the query result is to maintain the entire blockchain database and perform the queries locally. Doing so incurs high storage and computational costs on the users, precluding this approach to be practically deployable on common light-weight devices (e.g., smartphones). To this end, this paper proposes $\pi$QLB, a query language for blockchain systems that ensures both confidentiality of query inputs and integrity of query results. Additionally, $\pi$QLB enables SQL-like queries over the blockchain data by introducing relational data semantics into the existing blockchain database. $\pi$QLB has applied the recent cryptography primitive, i.e., function secret sharing (FSS), to achieve confidentiality. To support integrity, we extend the traditional FSS setting in such a way that integrity of FSS results can be efficiently verified. Successful verification indicates absence of malicious behaviors on the servers, allowing the user to establish trust from the result. To the best of our knowledge, $\pi$QLB is the first query model designed for blockchain databases with support for confidentiality, integrity, and SQL-like queries.