Abstract
This paper presents a novel technique for multi-hop localization by using mobility (MLM) in mixed line-of-sight/non-line-of-sight (LOS/NLOS) environments. In MLM, nodes equipped with ultra-wideband (UWB) radio first move randomly for collecting time-of-arrival (TOA) measurements and then apply a modified biased Kalman filter (MBKF) designed for mitigating NLOS errors. NLOS bias in the measurements is further mitigated by using the shortest path distance (SPD) selection. Node positions are initialized by using multidimensional scaling (MDS) and then estimated by using iterative trilateration accompanied by an error accumulation management. Our simulation results demonstrate that the positioning accuracy of MLM outperforms previous methods in obstructed building environments with a limited anchor heard (AH).