Efficient index update for moving objects with future trajectories

Recently, more research has been conducted on moving object databases (MOD). Typically, there are three kinds of data for dynamic attributes in MOD, i.e., historical, current and future. Although many index structures have been developed for the former two types of data, there is not much work to deal with the future data. In particular, the problem of index update has not been addressed with efficient methods. This paper proposes a novel spatio-temporal index based on PMR quadtree, which is called Future Trajectory Quadtree (FT-Quadtree). FT-Quadtree adopts a trajectory segment shared structure and depicts an efficient update algorithm. The performance studies have shown that FT-Quadtree has superiority to the traditional one in index maintenance.