Performance Tests and Application Strategies of Spatial Indexes in the Embedded-GIS Data Management

Spatial indexes arranged by some sequence, are related to the positions, shapes of geographic entities and relations among them, which describe the outline info of object identifier (OID), minimum bounding rectangle (MBR) and pointers to storage location. Usually, the index situated between the spatial operating algorithms and the entities, enhances the efficiency of spatial operations by filtering and erasing irrelevant ones. Especially, for resource-constrained mobile terminals with embedded-GIS, it´s necessary to establish proper index strategies to solve the problem of limited memories conflicting with enormous data and the requirement of real-time processing. The paper not only stated the design and realization of the Seq-list, KD-tree, quad-tree and R-tree indexes carried on embedded-GIS, but also applied these to Compaq iPAQ H3800 conducting massive careful tests. Based on analyzing and contrasting test results of memories occupied, time spent and power consumed, the paper appraised the validity of above indexes and proposed corresponding strategies for different embedded-GIS application environments