A Coordinated Energy Management Method For 5G Base Station Using Multi-Agent Deep Deterministic Policy Gradient

The increasing operation expenses (OPEX) of 5G base stations (BS) necessitates the efficient operational management schemes, among which one main approach is to reduce its energy cost through energy-efficient on-site management. In this paper, we propose a novel energy management method for 5G BS aiming to reduce energy costs through peak-load shifting, which involves the coordinated management of batteries and air conditioners. The air conditioners are used to assist thermal management of batteries and precool indoor air, enabling reliable and efficient utilization of both the electricity and cooling storage systems in the BS. Moreover, considering the independent executions of batteries and air conditioners in a unified environment, the multi-agent deep deterministic policy gradient (MADDPG) method is employed due to its properties of centralized training and decentralized execution. The simulation results show that the proposed method can effectively reduce the energy cost of BS and outperform all the compared methods.