Geospatial Data Sharing Based On SOA

It is a hot issue how to share data and integrate applications quickly and conveniently in different spatial information systems, on the Internet without converting data formats, knowing data's position and their owners, directly visiting the database, and with the absence of API.Firstly, in the light of the status of the development, and the existing problems of the spatial data sharing service, traditional methods and difficulties of the spatial data sharing has been summed up.The Web Service provides a method of solving spatial data sharing in a higher level and broader scope. With this method, the providers of the Web Service only expose the services of data or the operations. And users need not know the internal structure of services, and they are implemented by which language, only use the services under the service note (WSDL), can integrate the services into their applications.The REST style is an internet hypermedia distributed system architecture. It constructs system from the view of all being- resources, by providing a clear separation of concerns, hiding the low-level implement of resources and communication mechanisms, reducing the complexity of the architecture, transferring self-described data—Representation in the interfaces of REST components defined uniformly, in the constraints of using client-server, non-state, cache, unity interfaces, stratified, on-demand code and so on.On such basis, a general and mixed architecture of spatial data sharing model has been brought up based on the Web Service and the REST. Referencing to the service request, this model deploys, composes and uses the loosely coupled components through distributed network with its unique abstract, loosely coupled and large-grained characteristics.Secondly, the model of the service oriented spatial data sharing architecture based on the Open GIS, and the model of the service oriented spatial data sharing architecture with REST style have been explained, described, analyzed and evaluated respectively. Under the Open GIS framework, spatial data sharing models based on the WMS, WFS, WCS, and WPS Specifications have been constructed and designed, and their processes of implement have been elaborated. And the service oriented spatial data sharing architecture have been constructed with REST style.Thirdly, The SOA ideology and architecture Pattern has been introduced. Its most important feature is the separation of the service interface and the realization, which links the different function units of an application - Service and Resource, through the definition of good interfaces and contract between services.Then, the mainstream digital earth platforms -Google Earth & World Wind have been compared with each other. After a brief presentation of the local data applications, the solutions for network data expansion applications based on the Google Earth and the World Wind have been brought up. Based on World Wind Java SDK and AJAX, the clients of spatial data sharing have been developed and implemented.Finally, a spatial data sharing experimental platform has been established mainly based on the Open Source software, which is also compatible to the commercial map servers, GeoMedia WebMap, ArcGIS Server, ArcIMS, MapServer, etc.The REST and Web Service have been loaded, and integrated loading tested in varieties of clients respectively. The test result shows that this model of architecture can solve the issues of the spatial data sharing effectively.In this paper, the innovations lay in that a service oriented spatial data sharing model and architecture based on OpenGIS has been brought up; a service oriented spatial data sharing model and architecture with REST style has been proposed; a service oriented spatial data sharing model and architecture integrated the WebService, REST, and digital earth platform has been put forward, which makes the concerns of providers and users of the spatial data change from the data to services and resources, and enables the spatial data sharing implemented at a higher level; the clients of the digital earth platform have been designed and developed. And the integration of the service oriented spatial data sharing has been tested. The achievements of this paper are allowed the researchers, developers, spatial data providers and ordinary users to try and learn from them, so as to promote the improving of the theoretical study of the spatial data sharing and the developing in depth.