UWB Anchors Self-Localization for Indoor Environments

Ultra-WideBand (UWB) is a well-known technology in the field of position localization since positioning with an error lower than 10 cm can usually be achieved. UWB is considered a valuable technology for service robotics since it allows the precise tracking of moving objects, e.g. UGVs and UAVs, in GPS denied zones and indoor environments. However, the setup for a UWB localization system is time-consuming and inefficient as each anchor’s position must be accurately measured to obtain good performance from the system. This process may be tricky or impossible in harsh environments. Also, it is prone to artificial errors that may compromise the performance of the whole system. In this work, an algorithm for the self-localization of UWB anchors is proposed wherein the positions of the anchors are estimated by freely moving a ground vehicle (GV) equipped with a UWB tag. The proposed method is inspired by a Simultaneous Localization and Mapping (SLAM) technique used by the robotics community. An Extended Kalman Filter (EKF) estimates the anchors’ position and improves the GV’s odometry. Experimental tests demonstrate that anchors’ positions are estimated with an error smaller than 20 cm, also in the presence of noisy distance measurements. Furthermore, EKF hugely improves GV’s odometry since the error in the position does not increase with time, allowing precise and reliable localization of the tag.