Sie befinden Sich nicht im Netzwerk der Universität Paderborn. Der Zugriff auf elektronische Ressourcen ist gegebenenfalls nur via VPN oder Shibboleth (DFN-AAI) möglich. mehr Informationen...
This paper proposes a visible light-based positioning (VLP) model for estimating an object's three-dimensional (3-D) parameters such as height and radius, in addition to location in an indoor environment. The model is built using neural networks (NN), trained by simulating numerous multiple-object scenarios in an indoor environment. It takes into account the shadowing effects so it can be implemented in a crowded environment with multiple obstacles. The proposed algorithm has numerous applications, such as positioning-assisted communication, suspicious object monitoring, and surveillance in IoT sensor networks. The error in location prediction by the proposed model is approximately 2.4 cm. The errors in estimating the height and radius of the objects using the proposed framework are observed to be 4.39 cm and 1.37 cm, respectively. These errors can be alleviated at the expense of extra hardware. We also analyse the effect of system parameters like grid size and number of obstacles on the VLP system performance.