Residual-Based Anchor Selection Algorithm for NLOS Mitigation in Cluttered Indoor Environments

The accuracy of an indoor positioning system (IPS) under nonline-of-sight (NLOS) conditions is an open research topic. Performance of the IPS is significantly degraded due to a positive bias in range measurements caused by the NLOS condition between tags and anchors. This study proposes an anchor selection algorithm based on residual error, referred to as residual-based anchor selection (RAS), to mitigate the NLOS effect and improve the positioning accuracy of IPS. The proposed RAS algorithm does not require any prior information about the environment and utilizes only the measured distances between tags and anchors at each time instance to eliminate NLOS measurements, so its positioning performance is always guaranteed in both static and dynamic obstacles in environments. An ultrawideband (UWB) sensor is employed to validate the proposed algorithm because of its advantages for an IPS, such as high accuracy ranging, mitigation of multipath effects, and extended measurement range. The effectiveness of the proposed method is addressed through simulation and experiment.