An Application-centred Resilient GNSS Position Estimation Algorithm Based on Positioning Environment Conditions Awareness

A traditional receiver-centered GNSS positioning model addresses the GNSS resilience development with utilization of standardized global error correction models. Consideration of a GNSS receiver as a black-box unit that delivers position, velocity, and timing (PNT) services, renders the GNSS position estimation process inflexible for GNSS application development. Here we propose a novel concept and algorithm for a GNSS position estimation that relies upon the awareness of the immediate positioning environment conditions. A mathematical and algorithmic frameworks of the proposed approach in GNSS position estimation are outlined in which a mobile unit serves the radio frequency (satellite signal reception, condition and digitization) and base-band (pseudorange measurements, and navigation message parsing) domains. The GNSS position estimation and positioning environment effects mitigation become the responsibility of the navigation domain integrated with the targeted GNSS application. The accurate description of the immediate real-time positioning environment (geomagnetic, ionospheric, tropospheric, multi-path, but also jamming and spoofing) conditions is either obtained in real-time from mobile unit sensors, or provided by trusted third parties. The GNSS application adapts accordingly the GNSS position estimation algorithm, and deploys the pseudorange error correction models for the real immediate positioning environment conditions scenario. The application-centered GNSS position estimation algorithm becomes focused on the provision of the Positioning, Navigation, and Timing (PNT) Quality of Service (QoS) scaled to the application needs, thus providing the more efficient mitigation of the positioning environment adverse effects while at the same time optimising computing and energy resources. An initial proof-of-principle performance assessment with a bespoke statistical learning-based environment condition model in the case of a rapidly developing short-term geomagnetic storm shows up to 92% mean positioning error reduction, and more than 50% reduction in the positioning error standard deviation.