Joint Cache Placement and Beamforming Design for Scalable Videos in C-RAN

This paper investigates the joint optimization of layered caching and beamforming design for scalable videos delivered by Dynamic Adaptive Streaming over HTTP in the multicell cloud radio access network. To transmit differential scalable videos, each video is divided into a base layer and a set of enhancement layers based on scalable video coding. Taking the layered property of scalable videos into account, we propose a layered caching scheme, i.e., multiple layers of different scalable videos are cached in the baseband unit pool and the cache-enabled cluster respectively, which alleviates the burden of backhaul links effectively. Since the quality of videos received by each user depends on the available bandwidth, the appropriate scalable videos are provided for users based on the transmission rates. To provide superior video quality for users, we aim to maximize the sum rate for users accessing the remote radio heads and regard the backhaul consumption as a quality of service constraint explicitly. Then, the weighted sum rate maximization problem is formulated, which can be converted into a quadratically constrained quadratic programming problem through the l0-norm approximation technique and the weighted minimum mean square error approach. Finally, simulation results demonstrate the superiority of the proposed scheme.