QoE Driven BS Clustering and Multicast Beamforming in Cache-Enabled C-RANs
Ruijin Sun, Ying Wang, Nan Sheng Cheng, Haibo Zhou, Xuemin Shen · 2018
Pre-caching popular videos at the local storage of base stations (BSs) can significantly alleviate the tremendous backhaul burden. In this paper, we consider a cache-enabled cloud radio access network (C-RAN) scenario, where multiple BSs cooperatively serve multiple users. Each BS has a local storage and connects to the central processor (CP) via a backhaul link. Since multiple users may simultaneously submit the same request, the multicasting is also exploited to further offload the wireless traffic. The joint BS clustering and beamforming are optimized to maximize the weighted sum quality of experience (QoE) subject to the transmission power constraint and the backhaul capacity constraint. To solve this mixed-integer nonlinear programming, we first equivalently reformulate it as a sparse beamforming problem. Then, the reweighted ℓ1-norm technique is adopted to approximate the non-convex backhaul constraint and the successive convex approximation (SCA) method is applied to deal with the non-convex QoE objective. Simulation results show that cache strategies have great impact on the QoE performance and our proposed scheme significantly outperforms the traditional rate maximization scheme.