A two-phase UWB-based positioning method for indoor non-line-of-sight environments

When positioning in indoor environments using Ultra-Wideband (UWB), Non-Line-of-Sight (NLOS) range measurements will degrade positioning accuracy if they are not solved properly. This paper first reviews the existing solutions, and a novel approach named the Two-Phase Target Positioning (TPTP) algorithm is proposed. This algorithm involves a coarse positioning phase followed by a refined positioning phase. In the coarse positioning phase, the residual weighting algorithm is modified and utilized for generating the coarse position estimate which is then used for identifying the NLOS range measurements. In the refinement phase, a joint constraint region is established to facilitate the generation of prior samples within the Sequential Monte Carlo (SMC) method framework. The Subtraction-Average-based Optimization (SABO) algorithm is employed to update samples and search for the optimal solution, ultimately achieving refined position estimation. Experimental results show the superiority of the TPTP algorithm over both classical and some state-of-the-art positioning algorithms in terms of positioning accuracy. Furthermore, the proposed positioning algorithm exhibits an affordable computational load for real-time applications.