TDOA Fingerprinting for Localization in Non-Line-of-Sight and Multipath Environments

Localization in urban areas with no direct line-of-sight between the target and anchor nodes turns out to be challenging for many well-known localization methods. In combination with multipath propagation of employed radio waves, the performance of the location estimates can be heavily deteriorated. In this paper we introduce a multilateration method that is able to cope with such conditions. Employing a discrete approach for the estimation of the emitter location, we show how prior knowledge of the environment, obtained from a ray tracer or ray launcher, can improve the system accuracy. The localization algorithm is based on time differences of arrival between the observing sensors. However, in non-line-of-sight these time differences change as opposed to common signal models that assume free-space propagation. This mismatch is resolved taking into account information of the propagation paths provided by the ray launcher. Simulation results demonstrate that the proposed method effectively mitigates the impact of non-line-of-sight and multipath propagation on the location accuracy.