Clustering is a property that is needed to improve query processing performance by locating logically related objects that are likely to be accessed together in the physically same page. Also, high storage utilization of the index structure keeps the height low so that the number of pages accessed by a query is small. Accordingly, for efficient processing of spatial queries, spatial access methods(SAM) must not only have the clustering property but also guarantee high storage utilization.
In this thesis, we first show that the center transformation technique, a category of SAMs, has the clustering property. We next use the MBR-MLGF, an index structure, to store the center transformed spatial objects to show that the index structure keeps high storage utilization. For performance evaluation, we compare the split patterns and storage utilizations of the center transformation technique with those of the corner transformation technique using the MBR-MLGF. The results indicate that center transformation technique has storage utilization improved by as much as 28 percents and accordingly provides better performance than corner transformation technique.