Evaluation of spatial database queries on two different spatial database systems
In this thesis an accurate performance test for the calculation of isochrones is given. For the generation of isochrones several queries are sent to spatial databases. The performance test stores the runtimes of every single query to get a better overview of the execution behavior of the different queries. Hence its possible to conclude about the query runtimes and to prove in what way they can be optimized. Several spatial databases are evaluated to see which performs fastest. One of the databases under test is SpatiaLite, which reflects changes in the isochrone algorithms directly, because the optimizer of this database has only limited capability. Also the size of the dataset has an influence on the runtime, therefore more datasets with different size are tested. The results show that not always the database with the best optmization techniques is the fastest one, for some queries SpatiaLite performed faster as the database system PostGIS. But for complex queries SpatiaLite cannot keep up with PostGIS.