PV Panel/Battery Sizing and Resource Allocation for Smart-Grid Powered C-RAN

In this paper, we investigate joint photovoltaic (PV) panel/battery sizing and resource allocation for smart-grid powered C-RAN. We aim to minimize the total system cost, including capital expenditure (CAPEX) for PV panel and battery installation and long-term operational expenditure (OPEX). A particle swarm optimization (PSO) based sizing algorithm is developed to determine the sizes of PV panels and batteries offline. Leveraging the stochastic dual-subgradient method, a distributed online resource allocation and energy trading (ORE) algorithm is proposed to make real-time decisions of beamforming, service schedule and energy management. The feasibility and asymptotic optimality of the proposed ORE are established by theoretical analysis based on the Lyapunov optimization.