Moving objects in a geo-DBMS: Structuring, indexing, querying and visualizing moving point objects in a geo-DBMS context

The main question of this research is: What is the potential and performance of a geo-DBMS to structure, index, query and visualize spatiotemporal points clouds of moving objects? Some researchers have been developing spatiatemporal data structures. These structures have some disadvantages. For instance, the model made by Vazirgiannis and Wolfson is especially made for road networks and anoter data model (developed by Wolfson) is relatively complicated and can be used for objects that move freely in space like aircrafts. Some approaches have been the disadvantage that they a lot of redundant storage. To overcome these disadvantages, a new approach is introduced. Thi smodel could be used for every purpose that deals with moving point object data (this makes it generic) and it does noet contain any redundant storage. An efficient indexing method makes quering of the data in a DBMS faster. For many query tpes, indexing methods are available. In moving point object cases, most of these indexin gmethods are based on the R-tree. Often it is not known in advance whih queries are going to be done on a data set and which structure and which indexing methods are going to be chosen. So it needs to be investigated which indexing methods give the fastes access to the data. The main principle of this generix model is choosing a base table, from which, by using (materialized) views, three other data representations (based n different geometric data types) easily could be derived. In this way a set of four data representations is available for quering. These four data types are 2D points, 3D points, 2D lines and 3D lines. In the 2D representations time is regarded as an attribute. From tbis set, many queries cold be formulated, like the objects speed, direction or acceleration. After the model is introduced, an efficient querying and indexing needs to be found. BEcause in Oracle 9i Spatial only the 2D and the 3D R-tree are implemented, the only way to manipulate the efficiency of accessing the data by the user is by formulating efficient queries. To demonstrate that this generic model is fast and flexible, two chase studies have been done. In the first case, the data has been collected in advance and being analyzed afterwards. So, the data set is static. The second case deals with real-time data. The main conclusion is that the generic data model for storing moving point data in a geo-DBMS is flexible and efficient. The data can be accessed in a fast way, depending on the type of query and the moethod used for indexing.