Return to Session E1 Next Abstract

Session E1: Cislunar, Lunar, and Martian Positioning, Navigation, and Timing 1

Onboard Navigation System Design for Altus-1 and Strategies for LCRNS SISE Compliance
Tara Mina, Georgia Institute of Technology; Jason Leonard, KinetX Aerospace; David Gaylor, Shaun Stewart, Intuitive Machines; John Christian, Georgia Institute of Technology
Location: Windsong 7-8
Date/Time: Wednesday, Sep. 16, 8:35 a.m.

INTRODUCTION
At the time of writing, the Artemis II lunar flyby mission is approaching its next scheduled launch window, planned for early April (NASA, 2026b). If successful, this historic mission will mark the first time that humans have ventured into deep space since 1972. The crew will make further history as the first woman, the first person of color, and the first non-American to enter cislunar space and fly around the Moon.
While remarkable in its own right, this mission represents just one critical step within the broader National Aeronautics and Space Administration (NASA) Artemis campaign. The Artemis program is a flagship inter-national collaboration, uniting space agencies and industry partners around the shared goal of establishing a sustainable human presence on the Moon (NASA, 2026a), which also serves as a strategic stepping stone for future missions to Mars and even deeper space. To facilitate a sustained presence, the lunar south pole is of particular strategic (and scientific) interest due to its unique lighting conditions, where ridges and peaks of nearly continuous sunlight offer reliable power generation and thermal stability, and regions of permanent shadows host large deposits of water ice (NASA, 2023).

With the growing number of crewed and robotic scientific exploration missions, as well as plans for infrastructure development in the lunar south pole region, reliable navigation and communication services are essential. To address these needs, NASA’s Space Communications and Navigation (SCaN) program developed the Lunar Communications Relay and Navigation Systems (LCRNS) project (NASA, 2025b) and selected Intuitive Machines as its commercial partner to design and deploy the first set of LCRNS-compliant satellites (NASA, 2024). In accordance with the LunaNet framework (Israel et al., 2020; NASA, 2025a), LCRNS will also broadcast the Augmented Forward Signal (AFS), a “GNSS-like” signal that delivers navigation and timing services to passive lunar users.

Upcoming Launch of LDN-1, the First Lunar Data Network (LDN) Satellite:
Intuitive Machines’ current design for the Lunar Data Network (LDN) consists of a terrestrial ground station network and five lunar satellites, with their planned orbits illustrated in Figure 1 (Brack et al., 2025). The constellation is designed to deliver Initial Operating Capability-C (IOC-C) LCRNS services, targeting coverage within 15 degrees latitude of the lunar south pole and altitudes of up to 200 km, while meeting a sufficient geometric dilution of precision (GDOP) of at most 6 (NASA, 2022).

The first satellite, LDN-1, is scheduled to launch later this year, in November 2026. To provide reliable navigation services for lunar users, the LDN spacecraft must maintain precise, real-time positioning, navigation, and timing (PNT). LDN-1 achieves this capability through an onboard precise orbit determination filter, initially presented in Mina et al. (2025b). Since then, several updates and refinements have been made to the navigation filter for LDN-1, accompanied by several operational design considerations to ensure that the spacecraft complies with stringent navigation service requirements outlined for LCRNS (NASA, 2025b).

Navigation Service Requirements: Signal-in-Space Error (SISE) and Availability:
For LDN-1 to achieve LCRNS-compliance, the broadcast AFS navigation signal must meet stringent measurement accuracy requirements, with a position and velocity signal-in-space error (SISE) of at most 13.43 m and 1.2 m/s (3sigma) at 10 seconds, respectively (NASA, 2022).

The LCRNS services requirements document (SRD) specifies that the IOC-C critical services must be available 98% of the time (NASA, 2022). LDN is only considered to be available and providing IOC-C service when the GDOP requirement of at most 6 is met, and the contributing satellites deliver the AFS within the LCRNS SISE requirements. Because the GDOP requirement was designed to be satisfied in a cost-efficient manner by the nominal, five-satellite LDN constellation--without necessitating additional on-orbit redundancy--this availability requirement in turn imposes a corresponding performance constraint on each spacecraft. Indeed, LDN-1 must meet the SISE requirement at least 98% of the time when visible from the IOC-C service volume.

For LDN-1 to meet the SISE and availability requirement for the IOC-C level of LCRNS service, key considerations must be examined not only for the onboard navigation observables and filter design, but also for the overall system and operations design of the satellite -- a critical aspect of LCRNS/LunaNet constellation design that, to the knowledge of the authors, few works examine in detail.

Focus of this Paper:
In this paper, we present an in-depth investigation on various system and operational design choices to ensure LDN-1 meets its SISE and availability requirements, while reducing its reliance on terrestrial support.

We present updates to the onboard navigation filter architecture for precise orbit determination of LDN-1 since our most recent paper (Mina et al., 2025b), along with our current estimated breakdown of the SISE budget, which consists of (1) state knowledge errors, (2) realization errors, and (3) calibration errors. Additionally, we discuss lessons learned from the Lunar GNSS Receiver Experiment (LuGRE) (Parker et al., 2025; Minetto et al., 2025)--the first GNSS data collection experiment in cislunar space and on the lunar surface--and describe how these results inform the design of the onboard navigation solution for LDN-1 and future LDN spacecraft.

Across a one-month simulation period, we rigorously evaluate the expected SISE contribution from not only state estimation errors, but also predictive errors that arise when generating the broadcast ephemeris message which lunar users reference at a future time. This performance evaluation also informs several key design considerations in the overall system operations design for LDN-1, including:
- judicious use of ground station support within resource and cost constraints,
- strategic planning of spacecraft attitude maintenance maneuvers for faster post-maneuver SISE recovery, and
- suitable selection of the validity period for the AFS ephemeris and clock parameters message, seeking to increase usability considerations while also maintaining SISE requirements.

Each design aspect of LDN-1 operations is discussed in greater detail in the following sections, including their associated design constraints that must be maintained while also aiming to meet stringent service requirements. The paper presents the detailed analyses and design conclusions for each component, in preparation for LDN-1’s upcoming launch later this year.

DESIGN OF THE LDN-1 POSITION, NAVIGATION, AND TMING (PNT) ARCHITECTURE
For precise orbit determination, LDN-1 is designed to primarily rely on weak-signal GNSS, supplemented with occasional terrestrial ground station support in order to satisfy SISE and availability requirements. This section first presents an overview of the current spacecraft schematic and navigation filter architecture, then the following section discusses the operational strategies explored to ensure LDN-1 meets the LCRNS standards (NASA, 2022).

While here we outline the key components of the current architecture, the paper shares further details about these updates to the navigation filter design since our most recent work (Mina et al., 2025b). The paper also presents the analysis and corresponding system and operational design decisions that ensure LDN-1 meets the LCRNS SISE and IOC-C service availability requirements.

LDN-1 Spacecraft and Hardware:
Designed in-house by Intuitive Machines, the current LDN-1 spacecraft bus schematic is illustrated in Figure 2. While this schematic is blurred in Figure 2 for preliminary distribution, the finalized schematic and its detailed image are included in the final paper. LDN-1 is equipped with a high-gain L1/E1 GNSS antenna and a candidate multi-constellation (GPS, Galileo) receiver. To process the GNSS measurements, LDN-1 hosts a candidate space-rated oven-controlled oscillator (OCXO) for stable onboard time-keeping.

Co-located with the weak-GNSS antenna, LDN-1 is fitted with a Direct-to-Earth (DTE) antenna for ground communications. Correspondingly, LDN-1 orients its -Z face toward Earth, allowing the antenna gimbals to maintain precise Earth-pointing for continuous GNSS signal tracking and telemetry reception. On the +X axis, the payload deck hosts the AFS transmitter and the OpNav camera, which maintain precise pointing of the lunar south pole.

Onboard Propagation Model:
This section provides an overview of the LDN-1 onboard propagation model, with detailed discussion available in our prior work (Mina et al., 2025b). For the onboard clock states, LDN-1 uses a three-state clock model with propagation uncertainty based on an in-depth characterization of the candidate OCXO clock model. For the spacecraft dynamical states, the onboard model fidelity is designed to limit the unmodeled forces to the 10e-11 km/s^2 level, thereby maintaining an error of within 0.3 meters over a 2-hour propagation period.

In the paper, we present the most up-to-date onboard propagation model for the LDN-1 mission. To maintain the selected propagation model fidelity, the onboard model includes an 18x18 gravity model using spherical harmonic coefficients from the Gravity Recovery and Interior Laboratory (GRAIL) model (Goossens et al., 2020), along with third-body perturbations from the Earth and Sun. Updates presented here also include refined modeling of solar radiation pressure (SRP), with the ballistic coefficient now represented as a state within the navigation filter, as well as the addition of lunar relativistic perturbations.

Navigation Observables:
Weak-GNSS measurements are processed onboard at a 1 Hz rate to reduce reliance on continuous terrestrial monitoring while satisfying the SISE requirement (NASA, 2022). The primary GNSS observables include pseudorange and time-differenced carrier phase (TDCP) measurements, whereas Doppler (or pseudorange rate) measurements are not included due to their inherent correlation with the TDCP measurements in the candidate weak-GNSS receiver. To mitigate significant ionospheric and plasmospheric delays (Parker et al., 2025; Iiyama and Gao, 2025) compared to the expected measurement noise, single-frequency GNSS measurements exclude signals below an Earth-altitude mask of at least 4000 km.

Due to their dependence on both the current and prior state, the GNSS TDCP measurements are integrated via a delayed-state filter framework (Mina et al., 2025a). A UDU-factorized formulation of this delayed-state filter is derived and implemented in flight software for improved numerical stability. Relativistic time conversions (Leonard et al., 2026) are also modeled to ensure accurate processing of the GNSS measurements in the onboard navigation filter. Details are discussed in greater depth in the paper.

In addition to GNSS, LDN-1 also processes optical navigation (OpNav) measurements onboard, providing additional position and velocity observability, particularly horizon-based OpNav (Christian, 2021). While OpNav is not a primary performance driver for meeting SISE requirements, it offers a contingent measurement source to support autonomous orbit determination during an unexpected data outage and provides redundancy for independent verification of the onboard navigation solution and measurement processing.

Periodically, terrestrial ground station observables also supplement the onboard navigation observables to assist with meeting SISE and service availability requirements. This constitutes one of several operational strategies evaluated for LDN-1, discussed in detail in the following section.

OPERATIONAL DESIGN CONSIDERATIONS AND STRATEGIES TO MEET SERVICE REQUIREMENTS
While the selection of onboard navigation observables and filter architecture are central to achieving LDN-1 compliance with LCRNS IOC-C service requirements, operational design decisions are also equally critical. This is particularly the case given the stringent LCRNS requirements for SISE and service availability.

Several operational decisions require in-depth investigation due to their direct impact on navigation performance, including (1) the extent and cadence of ground station support for LDN-1, (2) strategic planning of attitude maintenance maneuvers to ensure rapid SISE recovery, and (3) selection of the validity period for broadcast ephemeris and clock parameters. Each of these factors are discussed in greater depth in the following subsections.

Ground Station Support for LDN-1:
Precise two-way radiometric measurements from terrestrial stations are available to improve state knowledge for LDN-1 and support mission requirements. This capability is particularly valuable to facilitate faster recovery times during challenging scenarios, including (1) periods of degraded GNSS visibility or geometry and (2) attitude maintenance maneuvers (e.g., reaction wheel desaturation or yaw-flip events, as discussed in the next section). During these intervals, the spacecraft state uncertainty for LDN-1 is inflated, leading to degraded SISE performance and, ultimately, reduced service availability.

At the same time, access to these ground resources is limited and incurs significant cost and operational complexity. The two-way measurements are typically collected over a fixed satellite pass before the data is uplinked to LDN-1, introducing a delay in processing terrestrial observations onboard. Thus, a tradeoff exists when designing the extent of ground station support: ensuring reliable LCRNS service while also mitigating resource and operational burden. Accordingly, the paper presents design considerations on both the required frequency and duration of ground station updates, based on its quantified impact on SISE recovery and overall service availability.

Planning Attitude Maintenance Maneuvers for Rapid SISE Recovery:
As with many modern spacecraft, LDN-1 employs reaction wheel arrays that spin to create internal torque and autonomously control the spacecraft orientation. This enables LDN-1 to maintain pointing constraints, such as directing the AFS transmitter toward the lunar South pole. Over time, the reaction wheels can build up enough angular momentum to exceed their maximum storage capacity (i.e., reach their maximum rotational speed) and become “saturated.” Thus, to maintain controllability of the spacecraft attitude, LDN-1 must periodically perform a reaction wheel desaturation maneuver, during which thrusters are fired to counteract the torque caused by decelerating the reaction wheels.

Although these maneuvers are designed to produce negligible translational force, small unmodeled delta-v (i.e., velocity perturbation) errors may be introduced, and these perturbations can temporarily violate the stringent SISE requirements. The onboard GNSS measurements enable LDN-1 to autonomously recover from desaturation maneuver events and return to service. At the same time, our early studies found that the speed of this SISE recovery strongly depends on when the maneuver was performed, across the lunar cycle, partly due to variations on Earth-to-lunar-orbital-plane geometries.

While operational constraints limit how long the desaturation maneuvers can be deferred, we leverage the available flexibility within these constraints to strategically plan these maneuvers (particularly when the overall geometry is more favorable), thereby speeding up SISE recovery time and increasing service availability.

Selection of the Validity Period for Broadcast Ephemeris and Clock Parameters:
LDN-1 generates the AFS ephemeris and clock parameter data messages based on its estimated state knowledge, before broadcasting this data to users. As a result, the transmitted parameters inherently represent future predictions of the spacecraft ephemeris and clock states. While longer periods of validity reduce the required frequency to parse and update the LDN satellite parameters, they also increase position, velocity, and clock uncertainty as these states are predicted out further into the future. Accordingly, this paper analyzes the impact of the AFS message validity period and presents the rationale for its selection on LDN-1, balancing usability considerations with compliance to SISE requirements.

EXPERIMENTAL SETUP AND PERFORMANCE ANALYSIS
To characterize the performance of LDN-1’s onboard navigation filter and study the impact of the various system and operations design considerations, we develop a high-fidelity lunar simulation framework. This simulation incorporates realistic spacecraft and clock dynamics, weak-GNSS and OpNav observables, and state error uncertainty propagation through the delayed-state filter. System performance is then evaluated against the LCRNS IOC-C requirements, providing insight into how design and operational strategies influence SISE and service availability.

Simulation Setup:
Spacecraft motion is simulated using a 100x100 lunar gravity model, along with third-body effects from Earth, Sun, and Jupiter, solar radiation/thermal pressure forces, and apparent perturbations due to lunar relativistic effects.

The navigation filter incorporates a dynamical process noise covariance consistent with the magnitude of unmodeled accelerations which defined the fidelity of the reduced-order, onboard propagation model, as described in the previous section. In particular, the acceleration process noise is modeled by a power spectral density of (10e-11 km/s^2)^2 * dt in each axis, where dt denotes the forward propagation time, in seconds. The filter clock state process noise is consistent with our prior characterization of the candidate OCXO clock drift behavior (Mina et al., 2025b).

Recent internal studies have quantified the expected magnitudes of unmodeled delta-v perturbations for LDN-1 arising from desaturation maneuvers. Consistent with these results, this simulation incorporates a corresponding inflation of the velocity state error uncertainty for each scheduled desaturation maneuver. Terrestrial ground station measurement passes are also modeled, with measurement noise informed by the two-way link budget between the ground station and onboard DTE antenna.

Key Performance Metrics:
In this study, the performance metrics used to evaluate the system design align with those defined for the LCRNS IOC-C service requirements. In particular, we assess the predictive state error contribution to SISE over the simulation period, which is based on the forward predictions of the spacecraft’s dynamical and clock states and the AFS ephemeris message validity period. From this predictive SISE, we also evaluate the percent time that the LCRNS service is available, while LDN-1 is visible from the IOC-C service volume. Both metrics are consistent with the definitions in the Introduction and are used to characterize compliance with service requirements. The impact of a given ground support schedule is evaluated by quantifying the resulting increase in service availability relative to fully autonomous operations (e.g., due to faster SISE recovery following a maneuver or GNSS outage).

Performance Analyses:
We conduct three sets of analyses to investigate various aspects of LDN-1’s navigation system performance:
(1) Monte Carlo Analysis of Filter Covariance. Monte Carlo simulations are performed over 5 orbital periods in order to evaluate the consistency of state estimation errors with the filter’s covariance estimates, identifying any statistical discrepancies or biases. This analysis further verifies that the filter covariance accurately represents the state errors statistics and can be leveraged to reliably assess state knowledge errors to SISE over a longer simulation period.
(2) Long-Term LCRNS Service Performance Analysis. We characterize the SISE performance and system availability over a one-month simulation to capture the full lunar cycle and associated variations in orbital geometries. The start epoch of 2027 March 1, 00:00:00 UTC is selected for this analysis, since four distinct GNSS outage periods occur during the corresponding one-month simulation period. This scenario serves as the basis for evaluating LDN-1’s onboard navigation filter and evaluating various operational design strategies, including desaturation maneuver scheduling, ground station support, and ephemeris validity periods.
(3) SISE Recovery Due to GNSS Geometry Analysis. While the GNSS GDOP is consistently poor throughout the lunar satellite orbit, we observe that certain GNSS visibility conditions and orbital geometries lead to noticeably degraded performance and sluggish recovery from maneuvers. These conditions introduce periodic state error uncertainty variations over the one-month simulation period, during which the state knowledge and predictive errors exceed SISE allocations. This analysis presents an in-depth dilution of precision analysis and identifies orbital configurations associated with degraded versus resilient navigation performance. This study provides guidance for operational design, including time spans when limited ground station support is most beneficial and when necessary desaturation maneuvers can be scheduled to minimize impacts on service availability for LDN-1.

ACKNOWLEDGEMENTS
This work is sponsored by Intuitive Machines (IM).

REFERENCES
Brack, D., Roorda, T., Mathur, R., Gaylor, D., Hartigan, M., Ryden, G., Crenshaw, J., Volle, M., and Stewart, S. (2025). Preliminary design of the lunar data network constellation under operational considerations. In Proceedings of the AAS/AIAA Astrodynamics Specialist Conference, Boston, MA, USA.

Christian, J. A. (2021). A tutorial on horizon-based optical navigation and attitude determination with space imaging systems. IEEE Access, 9:19819–19853.

Goossens, S., Sabaka, T. J., Wieczorek, M. A., Neumann, G. A., Mazarico, E., Lemoine, F. G., Nicholas, J. B., Smith, D. E., and Zuber, M. T. (2020). High-resolution gravity field models from grail data and implications for models of the density structure of the moon’s crust. Journal of Geophysical Research: Planets, 125(2):e2019JE006086.

Iiyama, K. and Gao, G. (2025). Ionospheric and plasmaspheric delay characterization and mitigation methodologies for lunar terrestrial gnss receivers. In Proceedings of the 38th International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GNSS+ 2025), pages 3193–3210.

Israel, D. J., Mauldin, K. D., Roberts, C. J., Mitchell, J. W., Pulkkinen, A. A., La Vida, D. C., Johnson, M. A., Christe, S. D., and Gramling, C. J. (2020). LunaNet: a flexible and extensible lunar exploration communications and navigation infrastructure. In 2020 IEEE Aerospace Conference, pages 1–14. IEEE.

Leonard, J. M., Stewart, S. M., and Gaylor, D. E. (2026). Relativistic framework for disseminating lunar time for the Intuitive Machines LCRNS PNT constellation. In Proceedings of the ION PTTI 2026 Meeting, Anaheim, CA. Institute of Navigation.

Mina, T., Marinello, L., and Christian, J. (2025a). Remarks on stochastic cloning and delayed-state filtering. arXiv preprint arXiv:2508.21260.

Mina, T., Thrasher, A., Hartigan, M., Leonard, J., Stewart, S., Antreasian, P., Pipich, K., Brack, D., Gaylor, D., Bedford-Dillow, B., et al. (2025b). Preliminary navigation system design for the first lcrns satellite providing lunar pnt services. In Proceedings of the 38th International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GNSS+ 2025), pages 2909–2946.

Minetto, A., Zocca, S., Vouch, O., Nardin, A., Dovis, F., Musmeci, M., and Facchinetti, C. (2025). Lunart-q: The lugre quick navigation analysis and reporting tool. In Proceedings of the 38th International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GNSS+ 2025), pages 1377–1390.

NASA (2022). Lunar relay services requirements document (SRD). Technical Report ESC-LCRNS-REQ-0090, NASA. Revision B, Available: https://www.nasa.gov/wp-content/uploads/2025/ 08/lunar-relay-services-requirements-document-srd.pdf.

NASA (2023). Why Artemis Will Focus on the Lunar South Polar Region. Architecture concept review 2022 white paper, National Aeronautics and Space Administration. Accessed Feb 25, 2026.

NASA (2024). NASA selects lunar relay contractor for near space network services. https://www.nasa.gov/ news-release/nasa-selects-lunar-relay-contractor-for-near-space-network-services/. Accessed Feb 25, 2026.

NASA (2025a). LunaNet interoperability specification document version 5. Technical report, National Aeronautics and Space Administration.

NASA (2025b). Lunar communications relay and navigation systems (LCRNS). https://www.nasa.gov/ goddard/esc/lcrns/. Accessed Feb 25, 2026.

NASA (2026a). Artemis: Humans in space – NASA’s Moon-to-Mars exploration campaign. https://www. nasa.gov/humans-in-space/artemis/. Accessed Feb 25, 2026.

NASA (2026b). Artemis II: NASA’s first crewed lunar flyby in 50 years. https://www.nasa.gov/mission/ artemis-ii/. Accessed Feb 25, 2026.

Parker, J. J. K., Konitzer, L., Anderson, B., Ashman, B., Bauer, F. H., Esantsi, N., Heiges, C., McKim, S., Miller, J. J., Sanathanamurthy, S., Valencia, L., Dovis, F., Minetto, A., Nardin, A., Vouch, O., Zocca, S., Bernardi, F., Boschiero, M., Fantinato, S., Miotti, E., Pozzobon, O., Pulliero, M., Tedesco, S., Facchinetti, C., Musmeci, M., and Varacalli, G. (2025). Initial results of the lunar GNSS receiver experiment (LuGRE). Presentation at the ION GNSS+ 2026 Conference, Baltimore, MD. Conference presentation.

Return to Session E1 Next Abstract