RangeCIR: A UWB CIR Dataset for Range Error Mitigation and LoS/NLoS Classification under Human-Induced Occlusion

RangeCIR is a dataset of ultra-wideband (UWB) ranging measurements collected using SynchronicIT SFM10-DEV development boards based on the NXP Trimension OL23D0 UWB chipset. It is intended for research on non-line-of-sight (NLoS) error mitigation and channel condition classification in UWB-based indoor positioning systems. The dataset includes estimated distances, ground-truth distances, LoS/NLoS labels derived from an optical line-of-sight tracking system, and Channel Impulse Response (CIR) data. Each CIR record consists of 136 complex-valued samples. Measurements were collected in a controlled indoor environment (6 x 8 meters) with human-induced occlusion to model realistic NLoS conditions. See README.md for additional details and usage instructions. For questions or feedback, please contact Andrii Yaroshevych at [e-mail].