A Data Sorting And Entities Joining Algorithm In Relational Database Management Systems

In current information system environments, one of the key challenges is how to handle the information explosion efficiently. Relational Database Management System (RDBMs) as a foundation and kernel of most of information systems faces the challenge especially.Various entities in the RDBMs products are stored in disorderly manner, because the theoretical basis and the requirements of engineering practice. In RDBMs, for disorder collection of data, the most basic and core processing functions are to sort them directly. In this paper, base on deep research and analysis of the existing common data sorting methods, aim at units have huge amounts of data within the data value with a high data density of this feature, made using of digital non-comparative map data sorting algorithm to sort data. At the first, it has good performance. At the second, it is not recursive function calls in the whole operation process, so the algorithm has a fixed complexity, increasing of data size will not lead to an overall performance increase of non-linear decline. At the last, as a non-comparison algorithm, both value and storage location of each element doesn't lead to operational performance fluctuation.On this achieved basis, this article summarizes the characteristics of algebra and relationship between the algorithm and them. For data reading, storage organization, association operations, and other point of joining in RDBMS environments, the paper applies a new large entities joining algorithm. It holds four good characteristics. The first and the most important one is very good executing performance。The sconed is functional, it is not only to accomplish tables joining but also to achieve the result set grouping. The third, the algorithm doesn't rely on any stored data components. This will not only greatly reduce the workload on the future of engineering practice, but also cut down the overhead of maintaining these components. Finally, it has stable performance. As with the new sort algorithm, this paper has a fixed physical joining algorithm complexity, increase the amount of data will not lead to disproportionately decrease the performance. In summary, although the bitmap data structure is already known, it is still a broad application space and great potential, which is worth further analysis and mining.