مشخصات پژوهش

Visible Light Positioning (VLP) is a promising technology for indoor applications due to its high accuracy, immunity to electromagnetic interference, and the dual use of existing LED lighting infrastructure for both illumination and data transmission. The utilization of light-emitting diodes (LEDs) as transmitters within the VLP framework offers several benefits compared to traditional radio frequency (RF) beacons, such as enhanced spatial resolution, reduced interference with sensitive electronic equipment, and compliance with electromagnetic compatibility (EMC) standards. These characteristics render the system particularly advantageous for healthcare applications, where precise indoor localization is essential for tracking patients, staff, and medical assets while ensuring minimal disruption to critical equipment. In this research, we have introduced a robust and practical VLP system tailored to align with the operational and infrastructural constraints of hospital environments. Specifically, we developed a comprehensive channel model by constructing the power delay profile (PDP) of the received signal, enabling us to capture multipath characteristics and analyze signal propagation in complex indoor settings. To evaluate the reliability of the direct signal path, we assessed the presence of the line-of-sight (LoS) link by extracting the root mean square (RMS) delay spread, a critical parameter in identifying the dominance of multipath effects. Recognizing the challenges posed by dynamic obstructions such as human movement or mobile medical equipment that may block the LoS path, we proposed a novel Modified Deep Gaussian Process (MDGP) model. This model leverages a combination of multiple kernels to flexibly adapt to varying channel conditions and effectively capture nonlinearities in the spatial distribution of received signal features. The MDGP is particularly designed to maintain localization accuracy in non-line-of-sight (NLoS) scenarios, where conventional me