分布在不同节点的空间数据库通过P2P协作共同完成复杂的空间应用任务。提出一种基于P2P的空间数据操作分布的引擎设计:引擎根据空间应用任务的解析,在P2P网络中动态地选择最佳的执行引擎。通过P2P的消息和通信机制进行协作,共同完成复杂的空间计算和空间查询任务。P2P的执行机制避免了集中式执行引擎带来的网络拥塞和单点失效问题,提高了空间数据资源和计算资源协作的可靠性和可用性。
Peer-to-peer(P2P)computing has triggered extensive research on distributed computing. This paper presents a P2P based distributed cooperative computing technology and a prototype system implementation, which facilitates eornplex spatial application by collaborating spatial data resources distributed in different peers. The system can parse a spatial query and dynamically select best quality peers to execute the query. Peers coordinate to finish the query based on P2P message exchange and a transaction mechanism. Compared to centralized ones, the dynamical execution control performed in a P2P fashion has several benefits, such as avoiding network congestion and single point of failure, so to improve the availability and reliability of spatial data resources and computational resources.