| With the rapid development of Internet in recent years, people’s demand for video conferencing is growing. Traditional video conferencing systems need enterprises to purchase conference hardware equipment and server resources independently, then provide machine rooms to build it. However, with the complex process, high technical threshold, and the expensive cost, it is simply not feasible for minor enterprises. This paper will put cloud computing into video conferencing, design and implement cloud platform, which targets to provide minor enterprises video conferencing services. Therefore, it has greater technical and social significance.Physical separation of different functions and single function clustered make cloud platform scalable. The devices classified by function and abstraction from software address the problem in cloud devices’diversity, support smooth upgrade. Cloud resource allocation strategy based on minimum time-interval proposed effectively improve the utilization. Clustering modules solves the loss problem of cloud terminal status real-time monitoring, and improve the accuracy.To enhance the performance and service quality and reduce operating costs of video conferencing cloud platform and cloud service, we further optimize overall architecture design, cloud resource allocation, cloud status monitoring and cloud load balancing modules, and then make corresponding tests. The results verify that the optimization programs are effective. |