Feasibility of PSD-Based Indoor Positioning in Smoke-Filled Environments Using Building Luminaires

Abstract

This article investigates the feasibility of a visible-light-based indoor positioning system (IPS) using position-sensitive detectors (PSDs) and commercial luminaires in smoke-filled environments. A dataset comprising 285 measurements was collected over 79 min in a controlled test room, with visibility levels ranging from 3 to 300 m. The system was evaluated under varying smoke densities using visibility as the principal metric, calculated via the Beer–Lambert law and Koschmieder equation. Experimental results show that positioning accuracy degrades with increasing smoke levels, with errors remaining below 270mm even under extreme conditions (visibility ≈3 m). Notably, the system detects signals from emitters outside the nominal field of view (FoV) via multipath reflections, with received power increasing under moderate smoke due to enhanced diffusion. In a challenging scenario—receiver placed 3 m away, behind an obstacle, and under dense smoke—the system maintained signal detection, enabling reliable roomlevel localization. These findings demonstrate the robustness of PSD-based IPSs in adverse environments and highlight their potential for emergency response applications, such as firefighter tracking and evacuation support, where coarse localization is critical.