Optimizing Power Beacon Placement in WET Systems via Self-Organizing Maps

This paper proposes a novel solution that jointly optimizes PBs placement and energy beamforming, relying solely on statistical channel state information (SCSI). The proposed method integrates Self-Organizing Maps (SOMs) with K-means clustering, both unsupervised machine learning techniques, to derive a suboptimal yet effective placement of PBs. Simulation results demonstrate that the proposed strategy reduces the total system charging time by up to 87% compared to random PBs deployment and achieves a performance close to that of benchmarks assuming perfect CSI (PCSI). Furthermore, the results confirm that appropriate PBs placement enables the use of simpler hardware at the PB -- specifically, by reducing the number of antennas required -- without significant loss in system efficiency, making the proposed solution suitable for practical WET deployments.