Finding the Right Place: Sensor Placement for UWB Time Difference of Arrival Localization in Cluttered Indoor Environments

TL;DR

This paper presents an algorithm to optimize UWB radio placement for indoor localization in cluttered environments, significantly improving accuracy by accounting for obstacles and NLOS biases.

Contribution

It introduces a novel optimization method for UWB radio placement that minimizes localization error considering environmental obstacles and NLOS effects.

Findings

Achieved 47% RMSE reduction in 2D localization

Achieved 76% RMSE reduction in 3D localization

Optimized radio placement reduces the number of radios needed for target accuracy

Abstract

Ultra-wideband (UWB) time difference of arrival (TDOA)-based localization has recently emerged as a promising indoor positioning solution. However, in cluttered environments, both the UWB radio positions and the obstacle-induced non-line-of-sight (NLOS) measurement biases significantly impact the quality of the position estimate. Consequently, the placement of the UWB radios must be carefully designed to provide satisfactory localization accuracy for a region of interest. In this work, we propose a novel algorithm that optimizes the UWB radio positions for a pre-defined region of interest in the presence of obstacles. The mean-squared error (MSE) metric is used to formulate an optimization problem that balances the influence of the geometry of the radio positions and the NLOS effects. We further apply the proposed algorithm to compute a minimal number of UWB radios required for a desired localization accuracy and their corresponding positions. In a real-world cluttered environment, we show that the designed UWB radio placements provide 47% and 76% localization root-mean-squared error (RMSE) reduction in 2D and 3D experiments, respectively, when compared against trivial placements.