A Prototype Spatio-temporal Database Built on Top of Relational Database

A spatio-temporal database manages both space and time data, whose position or extent may change over time. It is widely used in Transportation Management Systems and Geographic Information Systems to answer queries for objects at a particular position during a period of time. In this paper, we build a spatio-temporal database prototype by implementing a space-filling curve and R-tree inside a relational database. In particular, we use a relational database as a storage medium, but construct a Morton R-tree inside the relational database for fast data retrieval. Experiments with large real-world dataset show that our approach outperforms an existing modern spatial database by 40% for individual queries. Under concurrent queries, our system dramatically outperforms the existing commercial spatial database by a factor of 6.47.