A multi-node collaborative and iterative UWB localisation algorithm for indoor complex environments

Ultra-wide-band (UWB) has high positioning accuracy in line of sight (LOS) scenarios. However, in indoor environments with large regions and severe none line of sight (NLOS), the number of localisable nodes and positioning accuracy are unsatisfactory. In indoor complex environments with large regions, three propagation models exist: LOS, NLOS with direct path (DP-NLOS), and NLOS with none direct path (NDP-NLOS). This paper considers the indoor propagation model and proposes a multi-node collaborative and iterative UWB localisation algorithm that uses a set of anchor nodes to locate a small subset of unlocalised nodes and utilises information from these nodes to localise the remaining unlocalised nodes. Additionally, the algorithm utilises variance information of localisation inaccuracies to iterate for more accurate coordinates. Ultimately, we improve the node locatable rate by 67.7%. The root mean square error (RMSE) of localisation is reduced by 0.1 to 0.9 meters in different regions.