Positioning Integrity via Uncertainty Quantification

The integrity and reliability of the positioning information are crucial for applications, including safety-critical ones, that rely on cellular network users' locations. Release 18 of the Third Generation Partnership Project (3GPP) broadens the notion of integrity, which was initially applied only to positioning based on Global Navigation Satellite System (GNSS), to include positioning based on Radio Access Technology (RAT). This paper presents a general 3GPP-compliant framework to assess 5G positioning integrity. The framework is based on uncertainty quantification and is applied to classical time-difference-of-arrival positioning algorithms and to more advanced positioning methods that rely on machine learning. If an integrity alert is raised, error mitigation techniques are also proposed to enable resilient positioning through the selection of reliable network nodes. Results obtained by using a 3GPP-compliant dataset of position estimates in an indoor open-office scenario demonstrate promising detection and resilience capabilities in the presence of integrity breaches.