Grid-based Moving Object Trajectory Hybrid Index Structure

In recent years,with the development of positioning technology and wireless sensors,a large number of wireless handheld devices and in-vehicle devices with positioning functions have been widely used,enabling many new applications to generate a large amount of motion information.These time-varying position information needs Managed in the database.Traditional database systems usually assume that the attribute values remain unchanged before the data is explicitly updated,and therefore it is difficult to efficiently manage the continuously changing dynamic position information.In addition,the traditional index structure tends to miss between the spatial query efficiency and the efficiency of the emerging trajectory data query.Here,this paper proposes a grid-based moving object trajectory index structure,and gives a variety of spatio-temporal query algorithm based on this structure,the main content is as follows:Firstly,a grid-based moving object trajectory index structure is proposed for specific scenes.Based on the design of the outer structure and the TB-tree inner layer structure of the mesh,the construction algorithm of the index structure is given.Secondly,for all kinds of query operations of historical trajectory,in order to improve its spatial query performance,the algorithm of precise query,region query,k nearest neighbor query and trajectory query under grid-based moving object trajectory index structure is proposed,and the corresponding algorithm is given.Algorithm analysis shows that the efficiency of the query under the grid-based moving object trajectory index structure.Finally,an experimental system is built for the above mixed index structure,the system design idea is given,and the functions and operation examples of each part of the system are introduced.Compare and analyze the performance of different index structures under the same data set size.The grid-based moving object trajectory index structure and the traditional TB tree index structure are constructed for data sets with different data sizes,and comparison performance tests of region-wide queries and k-nearest neighbor queries are performed.Experimental results show that this kind of grid-based moving objecttrajectory index structure has good performance in the query.