Research On User Capability Optimization Strategy Of Desktop Cloud Based On SBC Model

The desktop cloud platform is an embodiment of cloud computing. It deploys the clients’desktop or applications into a cloud data center, and provides a kind of thin client equipment which has the functions of inputting or displaying computer interface. Using desktop cloud, the clients can run the applications on the remote virtual machines. The desktop cloud platform can allocate system resource for clients with good flexibility. But the efficiency of applications which run on the desktop cloud platform is restricted by the limited resource of the system, as well as the support ability of the desktop cloud. Through analyzing the specific characteristics of the desktop cloud based on Server-Based Computing (SBC) model, the strategy how to estimate the system capability, and a principle of calculating the system’s maximum capability are givens. It is to solve the problem of how many users is the optimal capability of the desktop cloud by using the strategy. Based on the principle, an algorithm which can optimize the policy of system capability is presented. It’s named Users Capability Strategy algorithm (abbr. UCS) that can efficiently determine the most optimized system capability of the desktop cloud.The basic idea of UCS algorithm is to get a value first by using the principle of calculating the system’s maximum capability, and then check the satisfactory response times of so many clients. If the all clients’response times are smaller than a threshold value, so the number of clients run in this time is the most optimized capability of system. Otherwise, the numbers of clients is decremented until the all the clients’response time are in the threshold scope. The experiments show when the result is applied to the servers of the desktop cloud, the UCS algorithm is better than other common solutions of estimating performance.The feasibility of the UCS algorithm is tested and verified in the wide used call center system as a case study. A call center of the Guangxi Unicom Company has a desktop cloud platform based on SBC model. It has300clients run the applications to receive incoming calls in the same time by using the desktop cloud platform. The servers of the desktop cloud platform are often slow down by their insufficient system capability. It also brings the result that the applications are response slowly to the user interface. By using the UCS algorithm, the most optimized system capability of a server in the call center desktop cloud platform is estimated in the experiments. Through analyzing the experiment data, the most optimized capability value of the call center desktop cloud platform based on the SBC model by the UCS algorithm is achieved.