Comparison of TDoA Algorithm Dimension and Modified Kalman Filter for Ultra-wide Band Indoor Positioning System

The Ultra-wideband indoor positioning system (UWB IPS) is a position tracking system that utilizes UWB transceivers. This research explores one of the commonly used techniques in UWB IPS, known as time difference of arrival (TDoA), and addresses two key issues related to TDoA-based UWB IPS: the selection of the TDoA algorithm dimension and filter design, specifically the Kalman filter (KF). This paper presents two distinct experiments to investigate these aspects. The first experiment compares two TDoA algorithms—the Chan algorithm and the Newton-Raphson algorithm—across both two-dimensional and three-dimensional settings. The second experiment evaluates the performance of a modified Kalman filter (KF) for TDoA-based UWB IPS, comparing it with the conventional, unmodified KF. The results from the first experiment indicate that the three-dimensional TDoA algorithm is better suited for indoor applications where tags and anchors are not always at the same level. Conversely, the two-dimensional algorithm performs better when the tag is level with the anchors. In the second experiment, the result shows that the modified KF outperforms the conventional KF, exhibiting adaptive behavior, reduced overshoot during sudden changes in movement, and improved convergence to the reference.