Research On Data Queries And Processing Techniques In Moving Objects Databases

With the development of information retrieval and related processing techniques that promoted by wireless sensor networks, the Moving Objects Databases (MOD) became essential for many applications that range from navigation and efficient traffic management, through environmental monitoring and target tracking. Investigating techniques for efficient query processing, computing and transmission of moving objects data has been an important topic of studies in academic research field.Almost all the related applications are rely on some form of Location Based Services. Due to the imprecision of system, sensor energy saving and users privacy protection, the uncertainty is unavoidable in location and tracking systems. The uncertain is an inherent property of the moving objects data in many domains, which caused the research attention in uncertain of moving objects data. As for the historical continues K-nearest-neighbor queries, the main scope of research were focused on the optimization of index data structure, or the traverse algorithms, such as the selection of best-first or depth-first. Little work was done in promoting the efficiency of updating the intermediate results of queries. So the overall efficiency of neighbor queries cannot be optimized. The MODs in distributed network environment have become an important emerging area of research, especially for the data transmission and aggregation of queries results. The proposed Bora algorithm partial resolved the transmission and aggregation problem of range query result in grid-like distributed network environment. But for neighbor queries and concurrent queries, the efficiency can be further improved. Work of the thesis resolved the above issues, and some innovative contributions are enumerated as follows:1. The algorithms of uncertain range queries for Bead/Necklace Model are proposed. Focusing on the problem of efficient processing of spatiotemporal range queries for uncertain trajectories, where the uncertainty of the motion was modeled as a necklace - a sequence of beads, queries are solved by mapping to computational geometry problems in 2D plane. For different spatiotemporal query predicate, the corresponding processing algorithms for the refinements stage are presented. The pruning algorithm is implemented by offset query polygon and object point location queries. It significantly promoted the processing of system queries. With those bases, the location error parameter is introduced to improve the defects of Bead/Necklace model. And external ellipse is introduced to simplify the complex issues in revised bead model. The corresponding query processing and pruning algorithms are also revised and keep the efficiency of processing queries. And the preliminary research work are did on quantitative uncertain range queries for Bead/Necklace model.2. A beach-line update algorithm is provided to improve the efficiency of historical continuous K-nearest-neighbor queries. In the research of historical continuous K-nearest-neighbor queries, the phenomenon of beach line in iteratively update the neighbor structure is found. By adding flag in node structure of query results, the beach-line update algorithm is proposed, which ensured the necessary update operation and eliminated the redundancy in previous algorithms. The proposed algorithm efficiently improved the efficiency of neighbor queries.3. The schema of query results data aggregation and transmission in grid-like distributed network environment are proposed. In grid-like distributed settings of MODs, the basic Bora (Bresenham-based Overlay for Routing and Aggregation) Algorithm solved the problem of range query results data transmission and aggregation. But for neighbor queries and concurrent spatio-temporal range queries, there still have some space to improve. The defects of basic Bora algorithm are improved and the efficiency for neighbor queries results data transmission and aggregation is promoted. And One-Tree-Bora and Piggy-Bora schema are provided to improve the efficiency when common concurrent queries were rendered. The problems of connection re-establishing and data redundancy duplication in transmission are eliminated and the efficiency of data transmission and aggregation is further promoted.