Modeling And Indexing The Trajectories Of Moving Objects In Road Networks

As the mobile communication technology and GPS technology develops quickly, there are more and more applications that have to deal with the spatial objects. These objects will change their positions as time goes by. Now people are interested in how to trace and manage these moving objects. However, moving objects are often moving in constrained networks in real world, especially in the road networks. The traditional index technology cannot manage the moving objects effectively.Firstly, this paper describes related technology about spatial index and spatio-temporal index and analyzes their advantages and disadvantages. Then this paper discusses the edge-oriented and route-oriented road model.Secondly, upon our own GIS platform named X2Net, this paper studies the trajectory of moving objects and designs two spatio-temporal indexes. The first one can be applied in unconstrained space. It adopts the mobile trajectories store strategy to index the moving objects'trajectories. The second one can be applied in constrained road networks. It adopts the mobile function store strategy and route-oriented model to index the moving objects'trajectories. These two indexes can be used in windows query, time query, history trajectories query and so on.The map mapping or road bind is the basis of applying mobile function store strategy and route-oriented road network model. This paper makes use of the road topology relationship and improves the road bind algorithm by the time and spatial locality principle while moving.Finally, this paper tests the new road bind algorithm. The experiments show good performance. The paper also compares the disk space cost and time cost when storing the trajectories between these two models.The two spatio-temporal index and improved map mapping algorithm have been successfully applied in logistics project, traffic police mobile information system, intelligent transportation system and so on.