Environment-Aware Power Beacon Placement for RF-WET IoT Networks Using DDPG

This paper proposes a Deep Deterministic Policy Gradient (DDPG)-based method for optimizing the horizontal placement of power beacons (PBs) in RF-WET IoT networks. The model considers Rician fading, distance-dependent path loss, nonlinear energy harvesting, and environmental effects from temperature and wind speed. The goal is to maximize the number of IoT devices whose harvested energy exceeds a minimum threshold. To handle the resulting nonconvex continuous optimization problem, DDPG is used to learn a continuous PB positioning policy. Results show that the proposed method learns stable policies and improves the number of served devices, highlighting the potential of deep reinforcement learning for RF-WET optimization in IoT scenarios.