An Efficient Algorithm for Searching Nearest Objects in Spatial Database

Abstract

Querying the nearest objects of a given point is an important function in spatial database systems. The algorithm called RKV [9] solves the requirement incompletely because only one nearest object can be found. Besides, the performance of RKV is low because its depth-first search causes backtracking of re-accessing some visited nodes. Based on breadth-first search, we propose a complete and efficient algorithm called CC in this paper to provide an alternative for high performance. To verify that CCis better than RKV in performance, several experimems were conducted on the efficiency of these two algorithms. The experiment results indicatu that CC performs better than RKV does about one-and-half to four-and-half-fold as the number of data objects in the database is increased. The more the spatial objects in a database, the better the performance of CC compared with that of RKV.