Application Research Of Mobile Agent In The Distributed Query System As Middleware

In recent years,with the wider application of distributed systems and the development of informationization, distributed database has become fashion. Distributed data query is one of the most application in distributed database system, and query optimization algorithm is the key in query processing. There is a tremendous stream data communication in the process of distributed query. A query algorithm with excellent performance can effectually reduce communications stream produced in the course of query. Otherwise, if the performance of the algorithm is poor, it will be possible to produce large communications stream and bring about network jam in some time. So, designing an algorithm with excellent performance is the key to reduce communications stream.The distributed database, the processing of distributed query, query optimization algorithm was analysised and researched. According to analysis and research of W algorithm and PERF algorithm, a new query algorithm(WPERF+ algorithm) is applied, and is proved theoretically that has a better performance compared to the W algorithm and the PERF algorithm. It can effectually reduce communications stream produced in the course of query. The new algorithm is applied in the process of distributed query, to a certain extent, solved network jam because of large communications stream.Mobile Agent is a new network computation technology, it can improve effectively system's robustness. Mobile Agent technology is applied in the process of distributed query. Because it only need that the network which route of Mobile Agent is unimpeded. Therefore, Mobile Agent can reduce distributed query defeat because of network interrupt.A distributed database middleware system based on the existing database is developed. It well solved the problem of location transparency and database type transparency and database type transparency and effectively decreased network jam and disruption. And the system was developed with Java, which has a better cross-platform features and portability.Finally, "Data Mining"-- hot of research and application at preseant is easily discussed and application prospect of distributed query applied in data mining is expounded in this paper.