Font Size: a A A

Research On Algorithms Of Aggregate Queries In Spatio-temporal Databases

Posted on:2011-01-11Degree:DoctorType:Dissertation
Country:ChinaCandidate:H XuFull Text:PDF
GTID:1118330332968063Subject:Computer software and theory
Abstract/Summary:PDF Full Text Request
The appearance of spatio-temporal database make people more convenient to store and manage spatial objects and spatio-temporal objects, and analyze spatial data and spatio-temporal data. With the extensive research on techniques of mobile computing and wireless transmission, spatio-temporal database has more applications in daily life. Therefore various new types of spatio-temporal queries have arisen. Not alike previous queries that only consider single query object (e.g., nearest neighbor query), these new types of queries take multiple query objects into account which are called spatio-temporal aggregate queries in term. Due to the complexity of spatio-temporal aggregate queries, it is urgent to give out efficient solutions to those queries. Since the querying algorithm directly decides the efficiency of query, it is important to find algorithms with high efficiencies for spatio-temporal aggregate queries.Group Nearest Neighbor query is a typical aggregate query in spatio-temporal database. It returns a data object such that the summed distance from this object to all query objects is minimized. Group Nearest Group query is the general form of group near-est neighbor and enriches the types of spatio-temporal aggregate queries. Group Nearest Group query returns a set of data objects such that the summed distance from these objects to their nearest query object is minimized. Since the number of candidate sets is huge, in order to reduce the searching space of data objects, pruning data objects that can't con-tribute to result according to current distance threshold and improving the quality of query result by iterations.The supporting of moving objects in spatio-temporal database makes it possible to monitor continuous queries over moving objects. Since moving objects change positions, speed or moving patterns frequently, it requires higher efficiencies of querying algorithms on moving objects. As the extension of Group Nearest Group query on moving objects, Continuous Group Nearest Group query focuses on how to efficiently monitor the change of results at every updating time. Utilizing the results of last update and local distance threshold can efficiently filter candidate data objects. Therefore the cost of frequent up-dates can be reduced. Querying results are improved during several iterations.As one of the supporting spatial objects (polygons) in spatio-temporal database, obstacles do exist in real life. The existence of obstacles make the visibility between two objects change. Hence the results of spatio-temporal aggregate queries may possibly be different. Moreover users may only interest in visible objects or querying result. Until now most queries considering obstacles as constraints focus on not multiple query objects but single one. Group Visible Nearest Neighbor query requires the returned object not only is visible to all query objects but also with the minimum summed distance to all query objects. Considering the visibility of query objects from data objects aspect can reduce the searching space of obstacles that have impact on query results. From query objects aspect, processing query objects as a whole and extending the visibility of single query object to multiple ones make it possible to prune both data objects and obstacles. Therefore the efficiency of querying algorithm is improved.
Keywords/Search Tags:Spatial Database, Spatio-temporal Database, Query Processing, Spatio-temporal Aggregate Queries, Nearest Neighbor Query, Moving Objects, Obstacles, Visibility
PDF Full Text Request
Related items