Hybrid Indexing of Moving Objects Based on Velocity Distribution

TPR*-tree is the most popular indexing method for the current and future position of moving objects,but its frequent updates performance is very low.In this paper a hybrid indexing method,HVTPR-tree,which is based on the velocity distribution is presented for moving objects with frequent updates.HVTPR-tree takes into account the distribution of both velocity domain and space domain.First the velocity domain is split regularly,and moving objects are classified into different velocity buckets by their velocities,thus the moving objects in one bucket have similar velocities.Then TPR-tree is used to index the moving objects in each bucket.To support frequent updates a supplemented hash index on leaf nodes is added to HVTPR-tree.Also an extended bottom-up update algorithm is developed for HVTPR-tree,thus having a good dynamic update performance and concurrency.Experimental results show that HVTPR-tree′s update and query performance outperform any other indexing methods including TPR*-tree.