Indexing the past, present and future positions of moving objects using PPFI*

Developing efficient index structures is an important issue for moving object database. Currently, most indexing methods of moving objects are focused on the past position, or the present and future one. Despite the fact that there are a few of indices to support query from the past to the future positions of moving objects, they either index broken “polylines” or have low query and update performance. In this paper, we propose a novel indexing technique named PPFI* which consists of a TB-tree and a HTPR*-tree. PPFI* not only supports queries of the positions of moving objects at all points in time, but also supports frequent update. Experimental results show that the update performance of PPFI* is better than that of the RPPF-Tree, and predictive query performance of the PPFI* are significantly improved compared with those of the RPPF-Tree.