| Experimental teaching is indispensable to higher education. It trains students' ability to apply theory to practice, and improves their innovation thinking and skill. However, there is a shortage of teaching experimental resources, both experimental rooms and experimental equipments, in colleges because the total budget for the education resources is limited in China. Thus, online experimental teaching applications driven by the information technology have become good examples that combined scientific research with teaching, and also have been approaches to cover the shortage of experimental resources. Up to the present, the work at home focusing on releasing the teaching-experiment devices, especially the real-experiment devices to the network as sharing resources for the cross-subject and cross-department research, has just started. Concluded from the development of technologies used by international online experiments, the key for the share of experiment resources is to incorporate distributed heterogonous ones into a platform for the interaction with users.After researching the latest progresses of remote experiment platform, this paper proposes a scheme of online experiment platform based on service oriented architecture which is a framework for accessing and integrating the experiments on Internet or Intranet. The model is realized with WCF on .NET platform, along with the service contract and data contract defined among nodes and the practicing method for realizing this platform introduced.According to the online experiment platform based on SOA, taking the a signal averaging experiment which is common in information science as an example, a prototype of online experiment system is introduced in this paper. In this prototype, the application of WCF to the experimental platform is practiced as well as that of other technologies. It is verified in the final part that the realized system is not only efficient in functions but have better load performance than those employing common Web service technology, by executing load testing experiment on the model system to obtain its performance-related data of experimental service. It is also concluded that the introduction of resource scheduling is useful in enhancing the efficiency of the whole online experimental platform. |