Otimização de Sistemas WET baseada na Clusterização de Dispositivos IoT

Marcelo Satyro, Victoria Souto, Samuel Montejo-Sánchez, Richard Demo Souza

DOI: 10.14209/sbrt.2024.1571029898

Evento: XLII Simpósio Brasileiro de Telecomunicações e Processamento de Sinais (SBrT2024)

Keywords: WET Beamforming Clusterização Aprendizado de Máquina

Abstract

Wireless energy transfer (WET) technology is fundamental to future wireless networks, as it can enable the uninterrupted operation of future Internet of Things (IoT) networks. Therefore, in this work, we proposed a novel approach to minimize the total recharge time based on the unsupervised machine learning technique K-means. Furthermore, since perfect knowledge of the channel state information (CSI) is a challenge for WET systems and is not available in most practical systems, in this work, we consider the design of the beamforming at the PB based only on the statistical CSI (SCSI) knowledge at the PB. Finally, we verified that the proposed approach based on a clustering strategy can reduce the total system recharge time by up to 75%. Furthermore, we demonstrate that the proposed solution achieves a close-to-optimal performance for different scenarios, demonstrating its robustness.

Download