Abstract
This study provides a comprehensive outdoor ultra-wideband (UWB) dataset to examine the multipath effects in line-of-sight and non-line-of-sight (NLOS) environments for real-time localization. Specifically, the dataset comprises static and dynamic datasets designed to capture discrete multipaths affected by antenna height, obstructions, and time-varying environments. A static dataset varies the antenna height and distance to analyze the multipath interference on the received signal strength and ranging error with a UWB pair and walls to replicate NLOS environments. These measurements reveal ranging errors from ground reflections and Fresnel zone signal attenuations depending on the antenna height, which should be prevented in anchor deployment. The dynamic dataset includes IMU, GNSS, and UWB measurements collected using a unique mobile robot system under various anchor configurations. This dataset aims to facilitate the evaluation of localization methods against high-accuracy RTK-GNSS. Two representative methods, least-squares and an error-state Kalman filter, are provided to support localization improvements. These datasets have accelerated the study of real-time localization because these datasets are the novel datasets measured in various outdoor environments.