The Benefits of Cooperation Between the Cloud and Private Data Centers for Multi-Rate Video Streaming

Video streaming is one of the applications with the highest traffic on the Internet. This high traffic leads to a lot of workload on the video servers (data centers), and increases the energy consumption of the servers. Reducing energy consumption becomes more important in the case that the data centers use renewable energy. The cost of these servers changes over time, based on the availability of energy resources such as solar and wind power. As a result, in order to reduce the cost of the servers, we need to use a mechanism to reduce the load on the servers, especially during times of energy cost increases. One way to tackle this problem is to lease some storage clouds that work as proxies during these periods of time, and provide the users with the popular videos through these storage clouds. However, finding the proper time during which the videos should be downloaded, and the efficient amount of storage, is not straightforward. Moreover, the popularity of the videos might change over different times of day, and for different geographic locations. In this paper, we find the optimal cloud lease that results in the minimum cost. For this purpose, we model the problem as a linear programming optimization, which can be solved in polynomial time. We also propose an optimal solution for the case of multi-resolution video coding, in which different users can request and watch the videos with different qualities. Keywords—Video streaming, renewable energy, storage cloud, multi-resolution video, optimization, video-on-demand, network coding.