Blockchain-Oriented Query Capability Optimization

Aiming at the problems of limited query types and low processing performance of traditional full-node-oriented queries, this paper adds relational semantics to block data to support complex queries composed of basic relational operators such as selection, projection, and connection. The system only stores one block of data, so no additional storage costs are added. In addition, transactions belonging to the same table are distributed to multiple blocks according to the timeline of their occurrence, which significantly increases I/O overhead. This paper designs table-level bitmap indexes and hierarchical indexes to speed up data access and optimize query processing. And ultimately significantly reduce the response time of the query.