Index Techenologies For Real-Time Databases

In a real_time Database System , Data and transactions may have explicit timing constraints , and thus the correctness of the system relies not only on the logic results but also on the time when the logical results are produced . In order to make the high performance of RTDB be met , there are many key theories and technologies must be solved.Index is an efficient tool that can improve the DBS'excuting performance . ARTs_EDB needs the main-memory database system (MMDB)'s support. So the index's selecting must be influenced by the memory's fast visiting character and high usage. The SB-tree(sequential binary tree), based on the AVL-tree and B+-tree, takes both trees'advantages: flexible operation and high memory usage.Index is an important support for time and spatial performance of RTDB. ARTs_EDB insures that there is no I/O access during transaction running time. And the precondition is that it must load necessary data from disk into main memory. From overall system, it will affect the throughput and make a time performance bottleneck. The optimized algorithm, first querying in main memery index then in secondary index, can obviously increase loading efficiency.Time is a supernatural thing, it dominates every aspect in our life.Advanced applications needs that DBMS can manage the temporal data,that's temporal database (TDBMS). Compared with the traditional Database , The temporal database has several points such as huge datum ,full of unequal predication and temporal query . These bring us into a embarrassment in query optimization , In order to insure its high performance of query ,a new indexing techniques , named time index, is introduced.At the same time,its arithmetic of search,insert,delete is implemented too. At last , the time index is applied in the temporal database .