Research On Pyramid Tree Indexing Based On Multi-mapping

With the rapid development of medicine, CAD, geography, or molecular biology, multidimensional data processing has been the focus of considerable research activities. As the focus of multidimensional data processing, multidimensional index always is a hot spot database research. The traditional database management system is one dimension space index, but for the multidimensional application, another index is needed to accelerate the search.As a method mapping the N-dimension to one dimension, pyramid technique can make use of the advanced of the database management system based on one dimension space to process multidimensional data, while need not adjust the applications. The pyramid technique is based on a special partitioning strategy which is optimized for high-dimensional data, the performance of the pyramid technique does not deteriorate when processing range queries on data of higher dimensionality, but the point search and range search is not good. A new model based on multiple mapping to over come the pyramid technique disadvantage is given. The new model uses multiple mapping to raise the efficiency of complexity search in multidimensional data space.A multidimensional data index system has been given based on the idea of pyramid technique, which has the plug-in peculiarity. This system overcome the shortcoming of which the generic multi-dimensional data struct does not compatible to the traditional database's B+ tree index struct, and the applications based on present commercial database management system need not adjust too much.Based on the analysis of the pyramid technique, a multidimensional data search system is designed and is actualized. In this system, the experiment compared between pyramid techniques and based on multi-mapping pyramid technique is done, and the experimental result is analysis. As the result of experiment, the pyramid technique based on multi-mapping can effective accelerate the search processing, and can incorporate into the commercial database management system.