MLIM: A Multi-Level Integrity Monitoring Method for 5G Indoor Positioning System

In the context of smart cities, 5G indoor positioning systems are crucial for improving the quality of location-based services. These systems are composed of multi-point positioning and fingerprint-matching algorithms. However, due to the complexity of indoor environments, 5G indoor positioning systems face challenges in accuracy and robustness. Traditional integrity monitoring algorithms cannot meet the fault monitoring requirements of fusion positioning systems in non-Gaussian noise scenarios. This paper investigates a multi-level integrity monitoring (MLIM) method. Firstly, this multi-level feedback structure enables the algorithm to monitor algorithm output and observations simultaneously, identifying and isolating potential errors promptly. Secondly, the Density-based anomaly detection algorithm is introduced into the MLIM framework to monitor system faults in non-Gaussian noise environments effectively. It identifies abnormal data points by evaluating their local density, maintaining efficient fault detection performance when the noise model is unknown or changing. Finally, the paper tests the performance of the MLIM algorithm at different levels. Experiments demonstrate that the proposed MLIM algorithm significantly reduces the root mean square error and maximum error of 5G positioning systems, and guarantees accuracy and stability.