Wireless indoor source localization as an inverse problem: An optimization approach via ray tracing simulations

Details

Supervisors

Oestges, Claude ; Jacques, Laurent ; Eertmans, Jérome

Faculty

Ecole polytechnique de Louvain

Degree label

Master [120] : ingénieur civil en mathématiques appliquées, à finalité spécialisée

Abstract

Ray tracing has emerged as an increasingly accurate method for modeling the propagation of electromagnetic waves in complex environments. Given a virtual representation of a scene and a specified transmitter position, it enables the estimation of the received power at the receiver locations. This motivates the use of ray tracing to address a corresponding inverse problem, namely, wireless indoor source localization. Wireless indoor source localization refers to the problem of determining the position of a radio transmitter in an indoor environment, based on measurements collected at multiple receivers and knowledge of the environment. In this work, we formulate this task as an optimization problem. We aim to determine the transmitter position that induces, through ray tracing simulations, measurements closest to those observed at the receivers. We first analyze the objective function resulting from the ray tracing model. We then propose resolution strategies based on first-order methods or derivative-free optimization, and evaluate the feasibility and the limitations of this approach.