Practical challenges of network optimized stored video delivery

Abstract

The scope of this thesis is to investigate the challenges and their possible solutions of applying a theoretical model of optimized stored video delivery to an LTE network. Recent paradigm shifts in content consumption, caused by the ever increasing popularity of online social networks, have allowed researchers to predict future content requests. In addition, models that predict user mobility as well as user activity can be used to infer the capacity that users of radio access networks will have in the future. Our work tries to combine the above, in order to optimize the delivery of video content that is predicted to be consumed by a user, whose future channel capacity is perfectly known, in a way that is transparent to him and require as few radio resources as possible. Thus, both increasing the number of users that can be served in a cell and the quality of service that they enjoy. The simulations have shown that our solution achieves a good balance between robustness and low channel utilization, while significantly reducing the cost compared to the benchmark application.