Indexing Techniques Based On Olap Data Warehouse Research

Data warehouse index technology is a data warehouse query optimization method developing from the traditional database index technology. Data warehouse index technology mainly includes Tree Index and Bitmap Index two methods. This article has done some deep discussion and research works about main feature, application area and development prospect of two index structures theoretically founding the wide application of two index technologies.Tree Index structure is suitable for high-cardinality dimension attribute, R-Tree becomes mainstream data warehouse multi-dimension index structure, and its advantage includes simple structure, easy maintenance, wide-application area, its disadvantage covers that index subspace data superposes redundancy and leads to query paths invalidation, influencing the query efficiency. Quadtree organizes index structure by the way of super-plane, and poses the advantage of fast query speed, convenient insertion and deletion execution. QR-Tree index combines the advantage of Quadtree and R-Tree, not only serves the need of high storing efficiency, but also avoid of many invalidation query, in order to achieve good query capability. Hence, here we especially discuss a new data warehouse multi-dimension QR-Tree index, and analyze the features of this index, offer the execution algorithms of insertion, deletion, query operate, make a comparison with the operation performance of R-Tree. According to experiment result, QR-Tree index is an ideal multi-dimension index structure and will have a good application prospect.Bitmap Index is another important data warehouse index technology, and executes the simplifying binary logic operation in order to optimize the query. The main feature of Bitmap Index is high query efficiency, easy implement and good manipulates. Section-based Bitmap Index, as the mainstream Bitmap Index technology, lowers the space price of index storing, but add the number of scanning. In order to balance the contradiction of two factors, we could use the combination operation ofbinary logic optimizing calculation to improve the query optimization. This article makes some improvement to membership-query combination simplifying algorithms and enhances the algorithms efficiency to some extent. Verified by the experiment, the improvement of algorithms efficiently optimizes the procedure of query and achieves the expected goal.On the whole, the index technology of data warehouse is still a new topic, R-Tree index and Bitmap Index have respective advantage, compound index combining the feature of two indexes also posses good query performance and is applied in wide areas.