Evaluating ARAIM Techniques for GNSS and LEO-PNT Positioning: A Dedicated Test Bench Analysis

Abstract

The advent of Low Earth Orbit (LEO) Positioning, Navigation, and Timing (PNT) systems represents a significant leap forward in satellite navigation technology. Unlike traditional Global Navigation Satellite Systems (GNSS) which operate in Medium Earth Orbit (MEO), LEO-PNT systems leverage satellites in closer proximity to the Earth, providing enhanced accuracy, availability and potentially integrity. The Advanced Receiver Autonomous Integrity Monitoring (ARAIM) represents a significant advancement of the RAIM service, specifically engineered for the aviation sector to enhance safety in navigation by utilizing multi-constellation satellite systems. ARAIM's design accounts for various possible faults, which become increasingly likely when several satellite constellations and frequencies are in use. Although originally developed for aviation, ARAIM shows promise for other safety-critical industries, such as the automotive sector and unmanned aerial systems (UAS), where the ARAIM concept must be adapted to challenging environments characterized by signal obstruction, multipath interference, and signal reflections. Presently, ARAIM receiver solutions should not miss the opportunity to improve their performance by incorporating the burgeoning LEO-PNT technology, which is expected to deliver a plethora of additional navigation signals with heightened power and enhanced performance, including better frequency and geometry diversification. In fact, the augmentation of GNSS with LEO-PNT, referred to as LeGNSS, is recognized as a key technology that can enhance single-point positioning (SPP) accuracy and accelerate the convergence of precise point positioning (PPP) solutions. Furthermore, it can bolster robustness and fault tolerance in safety-critical applications. However, integrating LEO-PNT systems with ARAIM is a complex task. It demands meticulous design of system integrity services and their flawless integration with the existing GNSS infrastructure. As indicated by research, conventional RAIM approaches may become impractical when dealing with a large number of ranging sources and high fault rates, which can lead to a cascading effect of faults across layers in a multitier architecture. Furthermore, sequential estimators such as used in PPP require adaptations from the snapshot ARAIM baseline. This paper presents an initial step toward extending the ARAIM concept to LeGNSS receivers that support MEO and LEO navigation signals. The study focuses on an L-band interoperable solution whereby the LEO-PNT constellation transmits signals close to the primary GNSS bands, simplifying the receiver technology's extension to the LEO signals and expediting market readiness. Specifically, this research draws from baseline configurations inspired by the recent EC studies [3] and commercial LEO-PNT players (i.e. Xona [4]) , considering three different solutions: 1) a high LEO constellation at 1200 km altitude in a Walker Delta configuration across four inclination planes; 2) a low MEO constellation of satellites at 8000 km altitude with the same inclination; and 3) a combined high LEO-low MEO solution. The goal is to maximize the benefits of the combined GNSS and LEO constellations, improving SPP and PPP performance and developing methodologies to assess their integrity across various scenarios. To this end, the first generation of the LeGNSS end-to-end (E2E) test bench developed at the JRC has been utilized to generate GNSS and LEO raw measurements (pseudorange, Doppler, and carrier phase) to input into the ARAIM navigation algorithm. This environment includes an RFCS simulator for rapid prototyping and creating synthetic GNSS and LEO measurements, as well as supporting hardware-in-the-loop testing of target GNSS receivers. The primary output is a LeGNSS RINEX file containing GNSS and LEO raw measurements based on the E2E test bench configurations. For LEO augmentation, additional software modules and interfaces have been developed. A LEO-PNT constellation design tool has been created to generate a combined GNSS and LEO constellation file that is propagated through the RFCS orbital engine. Considering the need to analyse different configurations in terms of visibility and access to PPP corrections, a dedicated module post-processes the GNSS and LEO navigation data to include representative ephemeris and clock errors, which are crucial for accurate simulation of the Signal-in-Space Range Error (SISRE). These models are calibrated according to the latest knowledge of GNSS and LEO systems, with broadcast and high accuracy service (HAS) ephemeris and clock accuracy informed by existing literature, while the LEO propagation adopts a dedicated error model consistent with recent studies. All signal-in-space propagation delays have been modelled according to established methods in the RFCS block, with local and residual error calibrations for raw measurements based on expected performance benchmarks for GNSS and LEO receivers. For the validation of the ARAIM concept, more specifically the failure detection and exclusion component, the E2E simulator is equipped with an artificial threat generator, which can trigger appropriate pseudorange ramps within the RFCS engine and inject degradation and failure analogous to clock jumps. In this work, the developed simulation environment is detailed covering the three baseline scenario’s. The performance of the LeGNSS solution is tested for both SPP and PPP solutions. Furthermore, the ability of FDE and the integrity bounds provided by the extended ARAIM concept are tested under feared event conditions