High-Precision 3-D Indoor Localization Using Pseudo-Centroid AoA Estimation and Adaptive Channel Preference With URAs

Three-dimensional (3D) indoor localization has garnered significant interest due to its applications in smart healthcare, 3D storage system, stereo interaction, obstacle avoidance, and collision detection. To improve measurement accuracy in 3D indoor localization, this paper proposes a pseudo-centroid localization algorithm integrated with an adaptive channel preference mechanism, leveraging measurements from distributed uniform rectangular arrays (URAs). By grouping URAs into pairs, we compute candidate positions from angle of arrival (AoA) measurements while analytically modeling the bias induced by non-coherent phase errors. The final position is derived via a weighted centroid method. A novel metric combining signal-to-noise ratio (SNR) and AoA estimation error is proposed to dynamically select optimal channels. To validate the effectiveness of the proposed methods, we conducted numerical simulations and indoor positioning experiments using Bluetooth Low Energy (BLE) devices. The results demonstrate that the pseudo-centroid method is robust and achieves high-precision indoor positioning.