Indexing And Querying Trajectories Of Moving Objects For Road Network

With the development and application of GPS positioning, wireless sensor technology, as well as the popularity of location-enabled handheld, automotive equipment, location-based service of moving object is widely used. Location of moving object changes over time, the database need to process the data which is changed frequently.the spatio-temporal data.Traditional database is designed to manage the static data, so it can’t adapt to update frequently data and moving object database comes into being.In road network, moving object database intend to query and manage the location of moving object,database usually contains amount of spatio-temporal data,establishing an efficient indexing mechanism for the moving object is one of the most important content on database.On the basis of introduction for moving object database technology, analysis for network-oriented index structure, this article use discrete representation. Firstly the edge of the road network model would be created.Secondly we also take full notice of the condition for road network topology in order to lead into neighboring edge information.we propose a index structure of managing historical trajectory and in-time information, the IQN-Tree(Indexing and Querying trajectories of moving objects for road Network).The network model contains a collection of intersection and a collection of edge.The IQN-Tree contains the upper’s structure and lower’s structure.Upper structure is organized by hash table and2DR-Tree which is based on edge. two-dimensional space of the road network is divided into several sub space by2DR-Tree which in order to achieve a spatial index. Hash table stores edge ID and neighboring edge information. Data record is connected with leaf node by two-way pointer.hash table would be set until moving object enter this edge when updating. The sub structure includes hash table of moving object, linked list and2DR-Tree forest,2DR-Tree forest’s index space are organized by relative position of edge and time intervals.2DR-Tree achieve spatial index of the moving object location information.Improving the lower R-Tree partition method and introducing principle of monotonically increasing time dimension. According to three way of entered moving object, the lower R-Tree for each edge is subdivided into three R-Trees, each leaf node pointer points to the lower three R-Trees,hash table stores moving object ID and in-time information, dynamic linked list for each data item stores moving object information on the base of time stamp’s updating. By looking for neighboring edge of information, this article designs a new window query algorithm and elaborates its implementation process. According to experimental comparison on MON-Tree, IQN-Tree has a good performance on insertion, window query, track query and road query.