High-Precise Multi-Cell Indoor Visible Light Positioning Based on Machine Learning for Indoor Millimeter-Wave/Terahertz Communication Assistance

Millimeter wave/terahertz wireless communication technologies are essential for high-capacity and high-data-rate services in Beyond 5G (B5G) and future 6G mobile communication systems. Due to the short wavelength and narrow beamwidth of these signals, high precision three-dimensional (3D) positioning of user equipment (UE) is required. Traditional beam tracking methods rely on sending training symbols, which reduces the system throughput and efficiency. This paper proposes and experimentally demonstrates a multi-cell 3D indoor high-precision positioning technology based on visible light communication (VLC). Using a two-stage artificial neural network (ANN), we achieve a positioning error of less than 1 cm in multi-cell indoor environments with cell volumes of 1.5 × 1.5 × 0.4 m3 and 0.5 × 0.5 × 0.4 m3. The proposed method can be integrated into B5G and next-generation indoor millimeter wave/terahertz communication and positioning systems. To the best of our knowledge, this work achieves the highest resolution for VLC-based 3D positioning reported to date.