UWB-Radio Distance Measurements Error mitigation due to Tag Orientation in Localization Systems

Abstract

An ultra-wideband (UWB) localization system presents an alternative in a GPS-denied environment. However, an orientation-dependent error is induced in distance measurements with UWB modules. This error has already been studied, looking at received power and the channel response signal parameters. In this paper, the neural network (NN) method for correcting the orientation-induced distance error in a realtime localization system without the need to obtain additional parameters is presented. The verification of the NN model was made in the experimental setup with 12 anchors and a tag. Results show that RMSE decreased by 5cm for the measured distance between the anchors and the tag compared to the calibration method that did not include orientation information. The use of the NN model resulted in a 15 cm decrease in localization error compared to results without the NN model.