Co-Design of Energy Storage and Resource Management for Cell-Free Radio Access Networks

Cell-free radio access network (CF-RAN) has emerged to address escalating demands in future mobile communications. The growing energy expense of CF-RAN necessitates efficient resource management and the integration of renewable energy harvesting systems. While prior studies on renewable energy-enabled CF-RAN have mainly focused on antenna resource management and energy consumption of access points (APs), they have largely overlooked significant power usage and scheduling requirements of central processing unit (CPU) and joint photovoltaic (PV) panel/battery sizing. To bridge these gaps, this paper integrates CF-RAN with smart grid technology and explores the joint optimization of PV panel/battery sizing and resource allocation to minimize system cost, including both capital expenditure (CAPEX) for PV/battery installation and operational expenditure (OPEX) for long-term operations. Leveraging the stochastic dual-subgradient method, we develop an online energy trading and resource allocation (OER) algorithm to enable real-time decisions on beamforming, service scheduling, and energy management. The feasibility and asymptotic optimality of the proposed OER are rigorously established through Lyapunov optimization techniques. Building on this asymptotically optimal solution, particle swarm optimization (PSO) is further employed to optimize the sizes of PV panels and batteries beforehand. Simulation results demonstrate a 62.7% reduction in total system cost compared to benchmark approaches, while reducing computational complexity in sizing PV panels and batteries.