Efficient Index Update for Moving Objects with Future Trajectories

R Ding,XF Meng,Y Bai

DOI: https://doi.org/10.1109/dasfaa.2003.1192382

2003-01-01

Abstract:Recently, more research has been conducted onmoving object databases (MOD). Typically, there arethree kinds of data for dynamic attributes in MOD, i.e.,historical, current and future. Although many indexstructures have been developed for the former two typesof data, there is not much work to deal with the futuredata. In particular, the problem of index update has notbeen addressed with efficient methods. This paperproposes a novel spatio-temporal index based on PMRquadtree, which is called Future Trajectory Quadtree(FT-Quadtree). FT-Quadtree adopts a trajectorysegment shared structure and depicts an efficient updatealgorithm. The performance studies have shown thatFT-Quadtree has superiority to the traditional one inindex maintenance.