Research On The Indexing Scheme For Online Data Of Moving Object

Moving Objects Databases records the location information of diverse moving objects all the time. Users can query the past, present, and future positions of moving objects in it. Moving Objects Databases have a wide range of applications in intelligent traffic,navigation,geographic information and military systems, etc. Although Moving Objects Databases has great prospect of application, but researches on is still in the early stage, therefore, researches on moving objects index is of profound theoretical and practical significance.This paper starts from the relevant knowledge of spatio-temporal database, mainly describing concepts and theories of moving objects'in spatio-temporal database. And the systematic classification of it is explained. With 3DR-tree index structure as the starting point, it has improved its imperfections, such as inconsiderate of the particularity of the time dimension, only dealing with offline data, declining index performance for those long-term stationary object. Through the node splitting approach, spatio-temporal database reduces the index number of long-cube so as to improve the index capability of historical data. Also by the separate index method of historical data and online data, the implementation of online data index is achieved. Then a new index structure is formed. Finally, a set of virtual datas from data generator have been used to test the transformation of 3DR-tree index structure for performance analysis, and from both the size of the storage space and the number of visits indicators, the 3DR-tree and the HR-tree index structure are compared to proving the query efficiency.The main contributions of this paper are as follows:Firstly, by using node splitting, the volume of MBR is reduced,also the destiny of data set so as to improve performance of 3DR-tree. Secondly, sa the result of tree splitting, it can form two trees structure, which can realize the online data index. Thirdly, aiming at the new index structure, some operations such as inquiry, inserting, have been designed. Lastly, virtual data generator has been used to prove that the performance has been improved by 20% in period inquiry.