The development of geo-computation drives spatial data infrastructure to the next generation, which should be intelligent, un-separated, and on-demand. On-demand service is one of the most important parts in the next generation spatial data infrastructure; it makes it possible to customize data supplies. In this paper, architecture of on-demand service is proposed. In this architecture, user or client program purchases data order by an XML based data order description language, and the data order is processed by on-demand data engine which coordinates ODDO interpreter, adapted data sources and deployed data processing units. At last, based on Spatial Information Grid Runtime Environment (SIGRE), a prototype of on-demand data service is implemented. And test on the prototype shows on-demand service can not only serve more powerfully, but also reduce service cost by less data transfer.
展开▼
