Nearest Neighbor Algorithms Using xBR-Trees

One of the common queries in spatial databases is the (K) Nearest Neighbor Query that discovers the (K) closest objects to a query object. Processing of spatial queries, in most cases, is accomplished by indexing spatial data by an access method. In this paper, we present algorithms for Nearest Neighbor Queries using a disk based structure that belongs to the Quad tree family, the xBR-tree, that can be used for indexing large point datasets. We demonstrate performance results (I/O efficiency and execution time) of alternative Nearest Neighbor algorithms, using real datasets.