UWB-Based Two-Way Localization With Unknown Anchor Position

Typically, the utilization of Ultra-Wideband (UWB) technology for positioning necessitates precise anchor positions. Nevertheless, the acquisition of anchor positions in public facilities is frequently impeded, substantially limiting the practical utility of UWB positioning and leading to resource underutilization. Current research does not adequately address the challenge of achieving high-accuracy positioning without prior knowledge of anchor positions. In light of this, this letter proposes a UWB-based two-way localization approach that does not require pre-existing anchor position information. It enables a UWB reader to rapidly ascertain the relative positions of unknown anchors and concurrently resolve its own relative movement trajectory. This letter has carried out validation experiments on closed-loop path and random curve path. The findings demonstrate the integration of Edge Artificial Intelligence (EAI) technology remarkably enhances accuracy, achieving average trajectory error reductions of 56.9% on closed-loop path and 12.5% on random curve path compared to the raw data. Furthermore, the Root Mean Square Error (RMSE) for unknown anchor positioning is 10.7 cm and 33.3 cm in two experiments, with respective accuracy enhancements of 44.9% and 22.2%.