Dynamic Base Station Clustering and Beamforming for an Uplink SIMO Cloud Radio Access Network

Consider the throughput maximization problem for an uplink SIMO cloud radio access network (C-RAN), where the base stations (BSs) can be clustered dynamically to perform joint reception under some backhaul constraints. In this paper, it is formulated as a sparse optimization problem based on dynamic BS clustering and beamforming. Instead of solving this problem via a series of feasibility checking steps (bisection), we design an efficient algorithm here. After the reformulation based on the semidefinite relaxation (SDR) and the Charnes Copper transformation (CCT), the BS clusters as well as the beamformers can be determined by solving only two simple convex SDP problems. Then an additional debiasing operation is introduced so that the throughput performance can be further improved. The effectiveness and efficiency of the proposed algorithm are demonstrated by some numerical examples.