Distributed or Centralized: An Experimental Study on Spatial Database Systems for Processing Big Trajectory Data

The continuous development of GPS-enabled equipments and location-based services has brought new challenges to the storage and management of trajectory big data. Traditional database system such as PostgreSQL/PostGIS to cope with this demand. Meanwhile, distributed trajectory management systems based on NoSQL have attracted increasing attention from both industry and acdemia. However, to the best of our knowledge, there are not many reseach discussing the comparision of these systems. This paper aiming at filling this gap by comparing the performance of the representitive relational database component PostGIS and the distributed NoSQL data management system GeoMesa in various range query and multi concurrent access scenarios through experimental analysis. The experimental results show that GeoMesa outperforms the relational spatial database component PostGIS in the large number of access results and multi-user access scenarios.