A new X-ray pulsar navigation system for Deep Space Exploration

Project Information

DeepSpacePULSE

Grant agreement ID: 101189496

DOI

10.3030/101189496

Project closed

EC signature date 18 November 2024

Start date 1 January 2025

End date 30 June 2026

Funded under

• European Research Council (ERC)

Total cost

No data

EU contribution

€ 150 000,00

Coordinated by

AGENCIA ESTATAL CONSEJO SUPERIOR DE INVESTIGACIONES CIENTIFICAS
Spain

Periodic Reporting for period 1 - DeepSpacePULSE (A new X-ray pulsar navigation system for Deep Space Exploration)

Reporting period: 2025-01-01 to 2026-06-30

Summary of the context and overall objectives of the project

Current Guidance, Navigation and Control systems rely extensively on ground-based orbit determination and human-in-the-loop operations. Existing autonomous navigation technologies, including GNSS, optical navigation and radar/lidar, provide effective solutions in specific environments but do not offer a universal, scalable solution for long-duration deep-space navigation. DeepSpacePULSE addresses this challenge by exploiting X-ray pulsars as natural navigation beacons. Their periodic signals provide stable celestial references distributed throughout the Galaxy, potentially allowing a spacecraft to determine its state with respect to an inertial reference frame independently of its distance from Earth. The project was conceived to assess the feasibility of a compact, low-power and autonomous X-ray pulsar navigation (XNAV) unit and to identify an observing and processing strategy capable of maximising navigation accuracy, robustness and autonomy. The original concept targeted a system with a mass below approximately 10 kg and dimensions comparable to a microwave oven, representing a major reduction compared with previous experimental systems. The project builds upon previous demonstrations such as NICER/SEXTANT and XPNAV-1, but addresses an important gap between demonstrating pulsar navigation and developing a practical autonomous navigation architecture. Previous demonstrations achieved kilometre-scale positioning but relied on comparatively large instruments and/or substantial external support. DeepSpacePULSE therefore focused not only on detector performance, but on the combined optimisation of the pulsar catalogue, observation strategy, measurement processing, navigation algorithm and instrument characteristics. The expected impact is to move XNAV towards a realistic autonomous navigation capability in which the spacecraft can reduce its dependence on Earth for state estimation. Such a capability could decrease ground-segment requirements and operational turnaround times while increasing spacecraft resilience and mission flexibility.

Work performed from the beginning of the project to the end of the period covered by the report and main results achieved so far

A first major activity concerned the characterisation and selection of navigation pulsars. The project built on a catalogue of 14 X-ray pulsars originally developed within the PODIUM activity, comprising young bright pulsars and millisecond pulsars distributed across the sky. The work investigated not only their X-ray brightness but also their pulse characteristics and long-term rotational stability. This directly addressed the project objective of determining which pulsars can provide accurate navigation measurements while minimising the need for updated timing information from Earth. A particularly important achievement was the replacement of simplified analytical measurement-error assumptions with an observation-based measurement model. Previous XNAV studies commonly estimated measurement uncertainty through signal-to-noise expressions or Cramer-Rao lower bounds. DeepSpacePULSE instead processed observational data obtained by NASA's NICER mission and rescaled these observations to represent the performance expected from smaller navigation instruments. This enabled navigation measurement accuracy to be characterised as a realistic function of pulsar flux, observation duration and detector effective area, including effects contained in real observational data that are absent from idealised analytical noise models. The analysis was then extended beyond X-ray flux to the other physical and operational parameters determining whether a pulsar is actually useful for navigation. The project investigated pulsar-spacecraft geometry, demonstrating that different pulsar directions provide different sensitivity to the components of the spacecraft state and that appropriate combinations of pulsars are therefore required to obtain balanced three-dimensional position information. It also modelled pulsar visibility, including interruptions caused by the Sun and, for Earth-orbiting spacecraft, occultation by the Earth. A further central activity concerned long-term timing stability. The propagation of pulsar timing-model uncertainties was evaluated to determine how phase-prediction errors evolve when timing solutions are extrapolated without ground updates. This quantified the fundamental distinction between bright but rotationally unstable young pulsars and fainter, highly stable millisecond pulsars. These elements were incorporated into an end-to-end autonomous navigation architecture using an Extended Kalman Filter (EKF) for onboard spacecraft state estimation. This implements the type of performance simulator envisaged in the original work plan, in which pulsar selection, signal/measurement generation and navigation performance are assessed together. The resulting system was evaluated in two representative scenarios: a spacecraft on an Earth-to-Jupiter interplanetary transfer and a satellite in a 600-km Low Earth Orbit. With an assumed detector effective area of 200 cm2, pulsar combinations containing the Crab achieved position errors below approximately 20 km during the interplanetary transfer and 7 km in LEO. These simulations also revealed an important limitation that is critical for the DeepSpacePULSE concept. Although the Crab provides the best short-term navigation accuracy because of its very high X-ray flux, its timing-model uncertainty eventually causes the navigation filter to diverge after approximately 20 days in the absence of timing-model updates. Replacing it with more rotationally stable pulsars decreases instantaneous positioning accuracy but allows the filter to remain stable for much longer periods. In the interplanetary simulations, stable-pulsar configurations maintained the navigation solution throughout the 365-day simulation without timing-model updates, demonstrating quantitatively the trade-off between maximum accuracy and maximum autonomy.

Progress beyond the state of the art and expected potential impact (including the socio-economic impact and the wider societal implications of the project so far)

Earlier studies generally concentrated on individual aspects of the problem - particularly detector area, observation time, source brightness or theoretical measurement uncertainty - and frequently relied heavily on the Crab pulsar because of its exceptional X-ray flux. The project demonstrates that this approach is insufficient for designing a genuinely autonomous navigation system. The first important advance is the development of an observation-driven performance assessment. Measurement uncertainties were derived from actual NICER pulsar observations and rescaled to representative compact-detector configurations rather than being determined exclusively from analytical bounds. The second advance is the demonstration of a multi-criteria framework for pulsar selection. Pulsed X-ray flux, timing stability, sky geometry and mission-dependent visibility have been evaluated together and propagated to spacecraft navigation performance. The results show that these criteria cannot be optimised independently. A source providing excellent range precision may compromise long-term autonomy, while a less luminous but stable pulsar can produce a less accurate instantaneous solution while maintaining navigation convergence without ground intervention. A Crab-containing configuration provides the highest accuracy - below 7 km in LEO and below 20 km during the Earth-Jupiter transfer for a 200 cm2 effective area - but becomes unsuitable for extended autonomous operation without regular timing-model updates. Stable pulsar combinations provide lower accuracy but substantially greater long-term robustness. This result directly refines one of the original DeepSpacePULSE assumptions: achieving very long autonomous operation cannot be reduced simply to selecting the most stable pulsars; the resulting measurement precision and navigation geometry must simultaneously satisfy mission requirements. A third important result is therefore the identification of pulsar selection itself as a mission-design optimisation problem. The work demonstrates that the optimal catalogue depends on trajectory, detector characteristics, visibility constraints, required positioning accuracy and acceptable dependence on ground-based timing updates. This represents an important step from demonstrating the physical feasibility of XNAV towards engineering a scalable autonomous navigation architecture. Overall, DeepSpacePULSE has established and quantitatively demonstrated the key design principle for practical autonomous XNAV: navigation performance is governed not by a single "best" pulsar, but by a mission-specific compromise between photon statistics, timing predictability, source geometry and visibility. In this sense, the results provide a concrete bridge between the original proof-of-concept vision and the future demonstration of XNAV as an operational technology for increasingly autonomous interplanetary and deep-space missions.