Motion in a geographical database system

Motion is an important phenomenon in geographic data related to Intelligent Vehicle Highway System (IVHS). Motion creates a need for frequent updates to the location of moving objects. Search queries like collision detection leads to a sequence of related spatial queries in the presence of moving objects. We evaluate the performance of existing spatial access methods for processing periodic collision detection queries in presence of moving objects. Tree based spatial access methods tend to incur high overhead for balancing and maintaining the tree structure for high update rate due to motion. We propose a new access method, called MoBiLe Files to efficiently manage spatial queries on moving objects. We use homeomorphic functions to map the object locations to disk addresses for saving disk space and computation time. We carry out detailed simulation and analysis of the performance of the proposed methods.