Calculating data warehouse aggregates using range-encoded bitmap index

A data warehouse is a database consisting of huge amounts of data collected from different source databases of an organization over a long period of time. Warehouse data are used for analytical purposes to make accurate and timely decisions based on previously integrated facts. Data warehouse is accessed using different kinds of analytical queries. One of the most critical issues is that those queries be responded to quickly and accurately. The size and logical schema of data warehouse systems make it difficult to apply existing query optimizing techniques originally developed for traditional database systems. Indexes are data structures, which help to locate the specific records in the database with minimum number of disk accesses.; Bitmap indexing is a promising technique for data warehousing systems, but space for bitmap indexes is a major problem. This thesis proposes the use of range-encoded bitmap index to calculate aggregates. By using space optimal range-encoded bitmap index for range predicates and aggregates, the need of separate indexes for these operations can be eliminated. The range-encoded index is efficiently used for evaluating range predicates. We are proposing algorithm to evaluate aggregates with the same index that gives equal performance, which was previously achieved by storing a separate index for these operations. This will reduce the space requirements and maintenance overheads considerably without losing performance for aggregates. The proposed indexing scheme is easy to maintain and use the population ratio of 1's in a bitmap to decide if the bitmap has to be scanned from the disk.