A Parametric Study on Autonomous Integrity Monitoring Using Non-Gnss Signals.

Navigation integrity is a well-established concept in the aviation domain. Motivated by the increased research on autonomy across all transportation sectors, integrity algorithms are now being explored for new use cases and measurement types. The attempt generally is to pair Global Navigation Satellite Systems (GNSS) signals with an inertial measurement unit (IMU) and measurements from other sensors, such as a camera or signals of opportunity. Two conflicting aspects make this a very challenging problem. On the one hand, tighter protection levels are demanded than those currently available in aviation. On the other hand, these new types of measurements have been less well studied than those from GNSS signals. In addition, they are often received in challenging environments such as urban areas, making it significantly more challenging to bound the errors tails. In this paper, we attempt to identify how this gap between demanded performance and less reliable ranging measurements from fifth generation (5G) cellular networks can be closed. First, we show how to include measurements from non-GNSS sources (i.e., 5G) into a well-established integrity algorithm. We then present a parametric study showing the effect of varying key parameters in the integrity algorithm, such as the measurement's nominal distribution, the probabilities that this nominal distribution is not valid, as well as the continuity and integrity requirements. Our results provide guidance on acceptable values for the nominal measurement model, they indicate that fault rates have to be kept below 10– 4 and that a relaxation of the integrity requirement can result in significantly tighter protection levels.