Abstract
We consider the resource allocation and beamforming design for system utility maximization in a cloud radio access network (C-RAN) under given fronthaul rate constraints and quality of service (QoS) requirements at users. The considered problem is known to be NP-hard and thus it is difficult to obtain global optimum solutions. Under practical assumptions, we approximate the considered problem using a quasi-geometric-programming problem. By analyzing its optimality conditions, we propose a fast iterative algorithm with low complexity to solve it. We show that the proposed algorithm is locally convergent. Simulations are carried out to verify the proposed algorithm. Compared to using standard convex optimization software, the proposed algorithm can reduce the processing time to as low as 1/700 while achieving the same final results.