Previous Abstract Return to Session B3 Next Abstract

Session B3: Lunar Navigation and Time

Cislunar Navigation Technology Demonstrations on the CAPSTONE Mission
Michael Thompson, Alec Forsman, Sai Chikine, Brian C. Peters, Advanced Space; Todd Ely, Jet Propulsion Laboratory; Dana Sorensen, Space Dynamics Laboratory; Jeff Parker, Brad Cheetham, Advanced Space
Location: Beacon B
Date/Time: Wednesday, Jan. 26, 11:03 a.m.

Peer Reviewed

Peer Reviewed

The Cislunar Autonomous Positioning System Technology Operations and Navigation Experiment (CAPSTONE) mission is an upcoming lunar flight demonstration, with a targeted launch in early 2022. This mission will serve as a pathfinder for operating in a Near Rectilinear Halo Orbit (NRHO) and accelerate the availability of the peer-to-peer navigation capabilities provided by the Cislunar Autonomous Positioning System (CAPS). The NRHO utilized by CAPSTONE is the same NRHO identified by NASA as the operational orbit for the Lunar Gateway, which is now in development. Since NASA is using this orbit for the first time, Advanced Space was able to couple our astrodynamics and navigation expertise with our desire to demonstrate CAPS to offer a mission that can accomplish both objectives on a very small budget and in a highly accelerated timeline.
This project will build and fly a 12U CubeSat overseen by Advanced Space partnering with Tyvak Nano-Satellite Systems for spacecraft integration, with several navigation-centric technology packages onboard. These include a small S-band communications payload capable of crosslink ranging with the Lunar Reconnaissance Orbiter (LRO), a dedicated payload flight computer for the CAPS demonstration, a chip-scale atomic clock to provide a more stable timing source than the standard radio oscillator, and an imaging camera. Note that this S-band payload is in addition to the X-band Iris radio that will be used for telemetry, tracking, and command. This system will be launched as a primary on a Rocket Lab launch vehicle using their Photon upper stage to perform the trans-lunar injection. Upon launch, the spacecraft will traverse a highly efficient Ballistic Lunar Transfer (BLT) taking approximately four months before entering the NRHO. CAPSTONE will then perform a primary demonstration phase in the NRHO for six months followed by a twelve-month technology enhancement operations phase.
The Cislunar Autonomous Positioning System (CAPS), a navigation product developed via a NASA SBIR that recently concluded its Phase II period of performance, is Advanced Space’s solution to the lunar congestion predicted to dramatically increase in the coming years as government, commercial, and international interests bring more missions to the Moon. By flying this navigation capability on the CAPSTONE mission, it will rapidly raise the TRL, commercialization, and infusion of CAPS in preparation of the cislunar traffic to come.
CAPS leverages substantial research in the field of dynamical systems theory. Using results from this research, CAPS creates observability between two spacecraft using asymmetry in gravitational acceleration fields (as found in cislunar space). In contrast to typical space-based measurements between spacecraft in similar dynamic environments, where only relative states can be determined, CAPS derives absolute position and velocity information about two or more satellites using only inter-satellite range and range-rate tracking data. This technique has been shown to work through rigorous testing by Advanced Space and others.
A CAPS network could be set up with any number of cooperating spacecraft, referred to as nodes, in cislunar space, including spacecraft in halo orbits, NRHOs, and spacecraft in orbit around the Moon itself. Generally, CAPS networks work best when one or more nodes are in a spatial (non-planar) libration point orbit due to acceleration field asymmetry considerations. Along these lines, asymmetry (in the three-body problem rotating frame) is generally beneficial, as well as vertical extent – distance from the rotating X-Y plane yields more dynamical asymmetry.
In a 2-node CAPS scenario involving a halo orbiter and a lunar orbiter, navigation solution accuracy can be on the order of 10m in position once converged. Thus far in navigation simulations, convergent solutions have been demonstrated between low-lunar orbiters and halo orbiters, low-lunar orbiters and NRHO orbiters, and between two halo orbiters.
CAPS scenarios with more than two nodes are also effective when measurements between pairs of spacecraft are scheduled with considerations to dynamical systems theory and estimation theory. For example, with a “smart” scheduler that schedules passes such that the line-of-sight aligns with the direction of maximum instantaneous uncertainty, state errors on many nodes can be kept at reasonable values.
To demonstrate and accelerate the infusion of the CAPS, CAPSTONE will perform several crosslinks with the Lunar Reconnaissance Orbiter (LRO). The tracking passes will occur when CAPSTONE is near periapse, as LRO is in a polar, low lunar orbit. The availability of these passes depend on a number of factors, including each spacecraft's power, their relative distances, lunar occultations, and pointing constraints. These tracking passes will provide two-way, coherent range and Doppler measurements to the CAPS flight software onboard CAPSTONE. These two observables will be ingested into the CAPS navigation software in order to estimate the inertial orbits of both spacecraft. The flight software will demonstrate CAPS in flight, while also downlinking the CAPSTONE-LRO crosslink data to the ground for further refinement and development. The objective of this demonstration is to accelerate the infusion of autonomous spacecraft navigation, where such crosslink tracking may support the navigation needs of both spacecraft in the link.
The crosslink between CAPSTONE and LRO is an S-Band link used to obtain two-way coherent ranging and Doppler measurements between the two spacecraft. The CAPSTONE radio is a reverse band radio used to simulate what a ground station communicating with LRO would generate. The carrier wave is modulated with a ranging signal which is then sent to LRO where it is coherently turned around in a bent pipe fashion back to CAPSTONE. CAPSTONE simultaneously generates and receives the signal. It uses the range tones and Doppler shift to compute a range and range-rate measurement.
These two spacecraft are able to communicate with one another while the distance between them remains less than 14,000 km, the line of sight between the two spacecraft is un-occulted by the Moon, and while CAPSTONE is above the local horizon of LRO. Additionally, the two spacecraft must maintain visibility with one another’s antenna beam patterns throughout the tracking arc by either slewing the spacecraft or gimbaling the antenna. For these demonstrations, CAPSTONE will perform an attitude slew to track LRO, while LRO will maintain its nominal attitude and gimbal its high-gain antenna to track CASPTONE. LRO currently resides in a low lunar orbit with a period of roughly 90 minutes, while CAPSTONE’s much larger orbit approaches perilune approximately every 6.5 days. Given these orbit geometries, the two spacecraft are within range to perform the crosslink for approximately 9 hours for every CAPSTONE orbit.
In addition to the crosslink demonstration with LRO, CAPSTONE will leverage the enhanced oscillator stability provided by an onboard Microsemi SA.45s chip-scale atomic clock (CSAC) to demonstrate one-way uplink ranging and range-rate. One-way ranging in this manner is a highly desirable capability which can serve to decongest terrestrial navigation systems by enabling all spacecraft in view of a ground station to utilize the same ranging signal. In contrast with two-way observables commonly used for deep space navigation, one-way ranging accuracy is highly dependent on the absolute timing accuracy of the spacecraft clock. Range is computed through comparison of the ranging signal’s Earth-transmit time (ETT) with the time of reception (TOR) at the spacecraft. This presents a challenge since the phase offset of the spacecraft clock (with respect to the ground station clock) must also be estimated to produce measurement accuracies suitable for orbit determination. Range-rate is computed as a “pseudo range-rate” that computes the difference in calibrated total count phase measurements divided by a time difference. Received one-way measurements will be processed onboard the spacecraft using the CAPS navigation software. CAPSTONE can perform the one-way navigation experiment for any of its scheduled ground passes with the DSN. For this demonstration, CAPSTONE’s Iris radio will simultaneously extract one-way measurements and turn around the timing signal such that ground-based two-way measurements can also be computed during Deep Space Network (DSN) tracking passes. This will provide direct points of comparison for further investigation of spacecraft and algorithm design considerations. Based on pre-flight experimentation and analysis conducted at JPL in support of the CAPSTONE mission, using this one-way technique, the expected measurement noise (1-sigma) for range and range-rate due to the onboard CSAC oscillator stability are approximately 2.5 meters and 11 mm/s (for a 60 second count) respectively.
The one-way phase (used to create range-rate) and range formulated by the Iris radio are originally referenced to the Iris’s local oscillator, a Q-Tech TCXO, that has a specified Allan Deviation (AD) of < 1E-9 at one-second (for the CAPSTONE Iris, the one-second AD was measured at ~4E-11 and the drift at 2.4E-8/day). CSAC provides a 1-pps input to the Iris radio that can used to improve the accuracy of the data because of its superior drift characteristics, measured at < 3E-11/day. The process to compute the range and range rate is dominated by quantization error and the TCXO drift; however, on longer time scales (over 60 seconds for range rate and 30 seconds for range) combined with appropriate processing algorithms, the CSAC-derived range rate and range begin to have improved accuracy as compared to the native TCXO-derived data and yield the noise figures stated previously.
To demonstrate the expected performance of CAPSTONE technology demonstrations, in this paper the authors will include high-fidelity navigation simulations of both the CAPS crosslink with LRO and the CSAC-stabilized one-way observables. These simulations will utilize noise values and biases that are expected based on ground tests of CAPSTONE hardware. While each of these cases can be run independently, they can also be run jointly – simulating the performance of an onboard navigation solution that can utilize both crosslink and one-way uplink measurements. The use of both data types can help demonstrate the expected performance of the onboard CAPSTONE autonomous navigation system.
During NRHO operations, CAPSTONE will also be performing two-way radiometric measurements with the ground. These measurements can be processed in conjunction with the two experimental data types to understand the reconstructed performance that could be post-processed on the ground. This type of analysis can show the specific accuracy improvements that CAPS and CSAC data types can provide while supplementing standard two-way observables.
The navigation experiments onboard the CAPSTONE mission will push the boundary of what has been demonstrated in terms of autonomous cislunar navigation. The results that will be presented in this paper represent the expected performance of these navigation experiments by the CAPSTONE navigation and payload teams. The authors look forward to CAPSTONE’s upcoming launch and demonstrating experimental navigation results to the community.



Previous Abstract Return to Session B3 Next Abstract