| Summary: | 碩士 === 大同大學 === 資訊工程學系(所) === 101 === With the popularization of the internet, various relevant applications of cloud services have been established, especially in the field of education. A cloud-based classroom by combining the personalized cloud storage and remote virtual classroom technology has been developed to enhance the learning and research performance of students. It increases the flexibility of resource utilization in time and space, thus, to the system manager, the efforts for software installation and maintenance can be saved and the system utilization and deployment can also be improved. However, the high-speed server of the cloud classroom system must be operated all the time and large power consumption will be incurred. In this paper, by applying the grey theory, four points of the GM (1,1) modeling prediction method, investigating and analyzing different time series of historical data and the situation of the prediction of the number of users for the cloud classrooms, and using the reciprocal of the absolute average residual percentage of traditional grey prediction method, a prediction model has been proposed to predict and adjust the number of cloud servers required dynamically. The proposed scheme will provide the cloud classroom managers a reliable information for deciding to open an adequate number of virtual machines, so as to achieve the purpose of both energy saving and service quality improvement.
|
|---|
