An Advanced Mobility-Aware Algorithm for Joint Beamforming and Clustering in Heterogeneous Cloud Radio Access Network

Abstract

Heterogeneous Cloud Radio Access Networks (H-CRANs) are a promising cost-effective architecture for 5G system which incorporates the cloud computing into Heterogeneous Networks (HetNets). We consider in this work the joint beamforming and clustering (user-to-Remote Radio Head (RRH) association) issue for downlink H-CRAN to solve the sum-rate maximization problem under fronthaul link capacity and per-RRH power constraints. The main objective is to address the beamforming and user association process over time by taking into account the user mobility as a key factor to tune the solution's parameters. More precisely, based on the mobility profile of users (mainly velocity), we propose an advanced Mobility-Aware Beamforming and User Clustering (MABUC) algorithm which selects the best Channel State Information (CSI) feedback strategy and periodicity to achieve the targeted sum-rate performance while ensuring the minimum possible cost (complexity and CSI signaling). MABUC inherits the behavior of our previously proposed Hybrid algorithm which periodically activates dynamic and static clustering strategies to manage the allocation process over time. MABUC algorithm, however, takes into account the user mobility by using a CSI estimation model which can improve the algorithm performance compared to reference schemes. Our proposed algorithm has the benefit to meet the targeted sum-rate performance while being aware and adaptive to practical system constraints such as mobility, complexity and signaling costs.