Queue-Aware Joint Remote Radio Head Activation and Beamforming for Green Cloud Radio Access Networks

Abstract

The cloud radio access network (C-RAN) is an emerging network architecture that holds the promise of coping with the explosive growth of mobile wireless data. In this paper, by considering the stochastic traffic arrivals and time-varying channel conditions, we address the stochastic optimization of joint remote radio head (RRH) activation and linear beamforming to maintain delay performance and minimize average network power consumption in downlink slotted C-RAN. We first formulate the joint optimization as a group sparse beamforming problem. Based on Lyapunov optimization technique, the stochastic optimization problem is then transformed into a queue-aware joint RRH activation and beamforming problem, which can be greedily solved at each slot. Finally, a low-complexity stationary algorithm with guaranteed convergence and closed-form expressions is proposed. The proposed algorithm can be implemented in a parallel manner, thus it is highly scalable to a large-sized C-RAN. Extensive numerical simulations validate the effectiveness of the proposed algorithm.