基于VO和Gossip的网格空间信息服务发现模型

A Grid Geospatial Services Discovery Mechanism Based on VO and Gossip

针对现有集中式和分布式服务发现模型的不足,提出了一种基于OWS和小世界模型的混合服务发现算法,该算法以虚拟组织为单位进行扩散,虚拟组织之间采用Gossip消息扩展机制来进行服务,在虚拟组织内部则基于集中式的目录服务进行服务匹配。针对传统Gossip算法采用纯随机策略进行消息发送的缺点,依据小世界模型,在虚拟组织信息表中随机增加shortcut,改进了算法的效率。构建了模拟试验环境,验证了算法的可行性和效率。

The existing services discovery methods were analyzed, such as centralized and decentralized. Their insufficiency was pointed out and a new method based on Gossip and VO was advanced, service requests were dispatching among VOs, and service matching was making in the VO by the VO. In order to improve efficiency or random dispatch by the classic gossip algorithm, theoretical results from Gossip were utilized. In the improved algorithm, a few shortcuts were inserted into the VO information tables with some probabilities, and the average path length was reduced. Experimental results indicated the feasibility and efficiency of this algorithm.