NLOS Measurement Identification Based on TDOA in Mixed NLOS–LOS Environments

The article delves into the critical challenges faced by wireless sensor networks in IoT applications, particularly the impact of non-line-of-sight (NLOS) measurements on localization accuracy. It introduces a novel method for identifying NLOS measurements based on time-difference-of-arrival (TDOA) information, which is particularly advantageous in mixed NLOS–LOS environments. The proposed approach leverages the Maximum Likelihood (ML) estimation cost function to evaluate the likelihood of line-of-sight (LOS) for each base station, followed by a hypothesis testing method to identify NLOS base stations. This method ensures high NLOS measurement identification rates and low computational complexity, making it suitable for real-time applications. The article also includes extensive simulations and real-world experiments to validate the performance of the proposed method, demonstrating its superiority over existing algorithms in terms of localization accuracy and timeliness. Additionally, the article discusses the potential for combining NLOS identification with error suppression algorithms to enhance the robustness of localization systems in complex and severe NLOS environments.