Adaptive TOA Estimation with Imperfect LOS and NLOS Knowledge in UWB Positioning Systems

Abstract

In recent years, localization based on Ultra-Wide Band (UWB) systems has been widely used due to its good positioning performance. However, the system reliability can severely degrade in non-line-of-sight (NLOS) conditions since time of arrival (TOA) estimation is highly conditioned by the presence or absence of line of sight (LOS). This manuscript evaluates an adaptive TOA estimation scheme based on prior knowledge of LOS/NLOS for TDOA-based UWB positioning. Assuming previous identification of LOS/NLOS channels, two correlation-based approaches have been employed for TOA estimation. The adaptive scheme has been assessed in terms of horizontal position error for several LOS and NLOS scenarios. Additionally, the robustness of the adaptive proposal has been studied when employing noisy LOS/NLOS estimates. Simulation results show that the proposed flexible solution outperforms traditional positioning schemes in different channel conditions despite errors in the LOS/NLOS estimates.