부산대학교 도서관

The degradation of indoor ultra-wideband (UWB) ranging accuracy caused by the non-line-of-sight (NLOS) condition has drawn considerable attention in the academic community. The existing solutions for generalized scenarios face challenges such as high cost, limited computational power, and lack of sufficient high-quality data, thereby hindering effective resolutions. Analyzing ranging errors caused by human body impact has been an effective approach for specific indoor scenarios. However, the accuracy of existing models and error correction schemes must be improved. After analyzing the error generation mechanisms of indoor UWB ranging, it was successfully identified that the two dominant error factors under human body impact include the human body shadowing error and the reference zero wall (RZW) reflection error. This paper proposes a particle filtering correction method based on Gaussian and generalized extreme value (GEV) distribution models and a direct compensation method. The experimental results show that the proposed Gauss-GEV error distribution model is more consistent with the measured values under line-of-sight (LOS) and human body shadowing conditions. The particle filtering correction scheme based on this model reduces the maximum mean error below 25 cm for human body shadowing error. A direct compensation scheme proposed to correct RZW reflection error also shows remarkable improvement in ranging accuracy. The human body impact model and error correction scheme for indoor UWB ranging proposed in this paper have great potential for indoor ranging applications.