Towards trusted and efficient temporal relational databases

Abstract

This research is to develop a blockchain based database based on a relational database management system (RDBMS). The database guarantees authenticity of the stored records using immutable and verifiable signatures. The database supports query evaluation over the entire timeline of the transaction log. To support efficient query evaluation, we utilize materialized snapshots at selected timestamps. We propose an optimal snapshot algorithm to compute the best timestamps so that the overall query load can be evaluated with minimal computation.