Multiresolution and Semantic Caching: A Framework for Network-Based Spatial Applications
Sai Sun · The University of Queensland · 2006
Spatial data and spatial database systems have attracted a considerable amount of attention for more than three decades and have been utilized in various areas to deal with geometric, geographic and multi-dimensional data. The applications have now been extended to the network environment (referring to both the Web environment and the wireless environment in this thesis) to support map-based applications, to provide online customized spatial queries and to offer location-dependent services. The full potential of spatial applications, however, has yet to be achieved due to performance issues related to the large sizes and high complexity of spatial data. Furthermore, compared to stand-alone spatial applications and non-spatial network applications, network-based spatial applications are confronted with the following issues specific to the network environment: m m · by providing a satisfiable and useful answer before 100% of the query results have been transmitted to the client. In addition, a novel replacement strategy is designed for MSC which takes the query pattern, window size and resolution range of MSC regions into account as well as the moving direction and semantic distance. Furthermore, this strategy adopts `Partial Discard' to avoid the cache granularity problem. Our extensive experiments on a large and complex real spatial database demonstrate a significant performance improvement using our proposed ideas.