The R*-tree: an efficient and robust access method for points and rectangles

The R*-tree is a hierarchical spatial access method for indexing points and rectangles, developed as a robust improvement over Guttman's original R-tree. The R-tree builds its directory by heuristically minimizing the area of the enclosing rectangle at each inner node. Through numerous experiments in a standardized testbed spanning widely varying data, queries, and operations, the authors found that optimizing area alone is insufficient, and designed the R*-tree to jointly optimize the area, margin, and overlap of each enclosing rectangle in the directory.

In an exhaustive performance comparison on the same testbed, the R*-tree clearly outperformed existing R-tree variants, including Guttman's linear and quadratic R-trees and Greene's variant, across different query and operation types such as map overlay, and for both rectangles and multidimensional points. It is practically attractive because it efficiently supports point and spatial data simultaneously while adding only slightly to implementation cost. The R*-tree became one of the most widely used spatial indexing structures in database systems.