服务部署与发布绑定的基于P2P网络的Web服务发现机制  

A P2P Based Web Service Discovery Mechanism with Bounding Deployment and Publication

作　　者：陈德伟[1] 许斌[1] 蔡月茹[1] 李涓子[1] 

机构地区：[1]清华大学计算机科学与技术系,北京100084

出　　处：《计算机学报》2005年第4期615-626,共12页Chinese Journal of Computers

基　　金：国家自然科学基金(60443002)资助.

摘　　要：Web服务技术的兴起为Web的使用提出了新的计算模型,构建基于面向服务的架构(SOA)的应用是未来分布式计算的发展方向.在SOA所要解决的技术问题中,服务的自动发现是难点和关键.文章提出了一种基于P2P技术和语义网技术的服务发现机制,将服务的部署与服务的发布无缝地绑定在一起,在处理用户查询的时候,提出了二阶段查询方法和两层搜索机制,提高了查询的效率.该机制由于采用了P2P技术,使得服务提供者可以自由地随时加入、变更和退出,不会影响最终用户的查询效果.文章还提出使用服务成长期和服务僵死期这两个指标,对服务发现机制进行性能评价.Web Services constitute a new computing model for Web application. The application based on SOA is a promising trend of distributed computing. The key and most difficult problem in SOA is how to automatically discover the services according to the end users’ query accurately and quickly. A new P2P and Semantic Web based service discovery mechanism is discussed in this paper in which the deployment and publication of a Web Service are bound together. When web service manager deploys his services, the embedded toolkit will create web service description files automatically and put them into the service metadata repository of the Peer. Profiting from using P2P network to exchange the metadata, the service providers are permitted to add and modify and delete services freely without conferring others. When a query is submitted by a service consumer, a two steps querying and two layers searching methods are used to improve the performance. Key words based matching and indexing in the Group layer are used during the first step, and semantic based Goal Capability matching in the Peer layer is used during the second step of querying. Two metrics, service growth time and service death time are also introduced in this paper to evaluate the discovery performance.

关 键 词：WEB服务发现 语义WEB服务 服务部署 服务发布 P2P网络 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]