3-D Localization of Wi-Fi APs and BLE Beacons With Interfloor Path Loss in Multifloor Buildings

Abstract

In modern urban environments with increasing numbers of high-rise buildings, accurate three-dimensional (3D) positioning is essential. Fingerprint-based positioning systems are widely used in such environments but require labor-intensive pre-survey efforts to construct floor-by-floor radio maps. In contrast, range-based methods demand less manual effort, requiring only anchor node (AN) positions and site-specific ranging models. However, obtaining this information remains challenging in multi-floor environments where inter-floor signal attenuation differs significantly from same-floor propagation. This study introduces an automated site survey procedure that jointly estimates 3D AN coordinates and site-specific path loss parameters, including floor penetration loss, by integrating smartphone-collected received signal strength (RSS) measurements with inertial sensor data for pedestrian dead reckoning (PDR). To address insufficient Wi-Fi coverage in real buildings, heterogeneous wireless infrastructure comprising Wi-Fi access points (APs) and Bluetooth Low Energy (BLE) beacons is considered, leveraging their similar 2.4 GHz propagation characteristics for unified modeling. Empirical measurements study was conducted to analyze and compare the inter-floor path loss of signals from these ANs, and a cost function was formulated to jointly estimate the 3D AN coordinates and path loss parameters. Experimental evaluation in a real multi-story building demonstrated mean absolute localization errors of 2.60 m for all ANs (1.02 m for Wi-Fi APs) with 100% floor identification accuracy. Using the estimated AN locations and path loss models, the online range-based positioning achieves 2.51 m accuracy over a 330 m trajectory, outperforming conventional approaches by effectively utilizing inter-floor signals through site-specific path loss modeling. © 2025 IEEE.