DRL-Based Energy-Efficient Pilot and Data Power Control for Uplink Cell-Free Networks

This paper studies an energy-efficient joint pilot-and data-power control for uplink cell-free massive multiple-input multiple-output (CF-mMIMO) networks. We first formulate an uplink energy efficiency (EE) maximization problem under the minimum required uplink data rate constraint. The formulated problem is non-convex and therefore hard to solve using conventional optimization algorithms. To solve the problem, we propose a deep reinforcement learning (DRL)-based scheme. Specifically, we use proximal policy optimization (PPO) algorithm. Simulation results demonstrate that the proposed PPO-based scheme can significantly enhance the EE of the uplink CF-mMIMO networks.