Index selection engine for spatial database system

The latest mobile Geographical Information System (GIS) technology is useful to manage spatial components of various daily business projects in corporate databases. It is important to apply proper geographical analysis efficiently in a wireless application. However, one of the problems of wireless internet is system bottlenecks that can slow down data processing in Mobile GIS. Spatial data indexing is one of the methods to speed up spatial queries. The existing spatial data indexing can only change the index used. However, an indexing method is only better for some ranges of data and conditions. The objective of this research is to speed up access to spatial database system by using spatial index selection engine. This research introduces an index selection engine for spatial database system for every condition and range of data, on top of the basic index structure. The index selection engine, which is called QuadRtree Selection engine, uses a rule-based Knowledge Base Expert System (KBES) to select between R-tree and Quadtree spatial data indices. These spatial data indexing methods are the best spatial data indexing methods among many other existing spatial index methods for low-dimensional spatial data which have different advantages and disadvantages based on the condition of spatial data. The result of using the proposed method can save time up to 42.5% compared to not using this method.