Font Size: a A A

The Research On Several Key Issues Of Distributed Heterogeneous Data Interoperation

Posted on:2008-03-30Degree:MasterType:Thesis
Country:ChinaCandidate:G H DengFull Text:PDF
GTID:2178360218952976Subject:Computer application technology
Abstract/Summary:PDF Full Text Request
With the development of computer network, more and more users begin to have the requirement to access and process information distributed among different data sources , which realizes the integration and sharing of various systems . It becomes more critical that move the application of information technology from transaction processing to data interoperation under the Internet. This paper focuses on distributed heterogeneous data interoperation, in which such technologies as construction of interoperation model, query optimization based on characteristics of obtaining local information are studied carefully.The traditional data interoperation models, federated database system and mediator/wrapper system , are summarized in this paper first . However , if federated database system is taken as the only approach on data integration, the ability of data sharing and efficiency of interoperation can not be guaranteed , and autonomous ability can not be satisfied if we take mediator/wrapper system as the only approach. This paper presents a distributed heterogeneous data interoperation model HIM which combines the advantage of two traditional model and can fulfill the real world application requirements much better.This paper then analyses the organization and conception of query processing in distributed heterogeneous data(DHD) interoperation . According to the characteristics of how to obtain local source information , global optimization is classified as query optimization schemes based on static local information and dynamic local information. Global query optimization is improved as the following . Firstly, the concepts as query graph with redundant clause, boundary point and query block are introduced in traditional Kruskal algorithm and the improved algorithm MQB-K mainly consists of query graph dividing , query block optimizing and query graph reconstructing. Secondly , based on MQB-K , the algorithm is further improved through defining different weights in query optimization schemes on static local information and the multiple-weights static query algorithms is obtained . Finally, combining optimization schemes based on dynamic local information , a hybrid algorithm DS-HO is designed. At the end of this paper, the algorithm is realized by experiment. The simulations and experimental results show that the performance of MQB-K is better than Kruskal and the hybrid optimization scheme DS-HO is much better than query optimization schemes based on static local information and on dynamic local information .
Keywords/Search Tags:distributed heterogeneous data(DHD), interoperation model, query optimization, weight function, query graph, local source information
PDF Full Text Request
Related items