AAS 06-223
THE JPL ROADMAP FOR DEEP SPACE NAVIGATION
Tomas J. Martin-Mur*, Douglas S. Abraham†, David Berry‡,
Shyam Bhaskaran#, Robert J. Cesarone&, Lincoln Wood§
This paper reviews the tentative set of deep space missions that will be
supported by NASA’s Deep Space Mission System in the next twenty-five
years, and extracts the driving set of navigation capabilities that these missions
will require. There will be many challenges including the support of new
mission navigation approaches such as formation flying and rendezvous in deep
space, low-energy and low-thrust orbit transfers, precise landing and ascent
vehicles, and autonomous navigation. Innovative strategies and approaches will
be needed to develop and field advanced navigation capabilities.
INTRODUCTION
The NASA Deep Space Mission System (DSMS) comprises both the Deep Space Network
(DSN) antennas, with the associated communication and control equipment, and the Advanced
Multi-Mission Operations System (AMMOS). The AMMOS provides tools, products and
services to help operate NASA’s missions, including those required to design and navigate
missions.
DSMS periodically compiles the set of future missions that it will be called to support for the
next twenty-five years, in order to forecast which communications and navigation capabilities
will be required by those missions. DSMS uses this mission set to study and plan the evolution of
the DSN and the AMMOS, and prepares a roadmap for the next twenty-five years. The DSMS
roadmap lists strategic goals mostly in relation to communication capabilities, but also includes
the improvement of spacecraft tracking and navigation services to both enhance current
capabilities and to enable new classes of future missions.
FUTURE MISSION SET, CHALLENGES AND REQUIRED CAPABILITIES
The set of future NASA deep-space missions changes periodically, as missions are added to
or removed from the forecast. The set of missions in a five-year horizon is well known and fairly
stable, but trying to forecast up to twenty-five years from now requires some educated guesswork.
NASA produces a Strategic Plan, and a number of NASA panels and committees prepare
roadmaps and surveys for different themes such as Solar System Exploration, Robotic and
*
Navigation and Mission Design Element Manager for the Multi-Mission Ground Systems and Services Program
Office, Jet Propulsion Laboratory, California Institute of Technology, 4800 Oak Grove Drive, Pasadena, CA, 91109,
USA, tmur@jpl.nasa.gov
†
Strategic Forecasting Lead, Architecture and Strategic Planning Office, Interplanetary Network Directorate, Jet
Propulsion Laboratory, California Institute of Technology
‡
Senior Engineer, Navigation and Mission Design System Engineering Group, Guidance, Navigation and Control
Section, Jet Propulsion Laboratory, California Institute of Technology
#
Technical Group Supervisor, Outer Planets Navigation Group Manager, Guidance, Navigation and Control Section,
Jet Propulsion Laboratory, California Institute of Technology
&
System Program Manager, Architecture and Strategic Planning Office, Interplanetary Network Directorate, Jet
Propulsion Laboratory, California Institute of Technology
§
Program Manager, Guidance, Navigation, and Control Section, Jet Propulsion Laboratory, California Institute of
Technology
Human Exploration of Mars and others. From those sources DSMS compiles the set of missions1
that most probably it will need to support, coordinating with other NASA mission modeling
efforts, in order to ensure emergence of, and consistency with, a commonly agreed upon
“snapshot” of potential mission customers over the next twenty-five years. Recently, such
coordination has involved building a consensus view relative to the Space Communications
Architecture Working Group’s (SCAWG) Integrated Mission Set (IMS) and the Program
Analysis & Evaluation (PA&E) Office’s Advance Mission Planning Model (AMPM). Once the
mission set is defined, the DSMS strategic planners then research mission requirements
documents, review presentations, and concept studies to derive the communication and
navigation needs for each mission2. For future competed opportunities, such as Mars Scouts,
New Frontiers and Discovery, DSMS needs to make some educated guess of which are the most
probable mission types. In addition DSMS needs to think about what would be the needs of other
future missions that are not currently possible given the current DSMS infrastructure and
capabilities. Competed missions are not likely to ask for new DSMS capabilities, because doing
so may increase the cost and risk posture of the proposal, and therefore make it less likely to be
selected. However, it would be foolish for NASA to always constrain itself to its current DSMS
capabilities, because doing so would exclude many worthwhile projects, and asymptotically
decrease the exploration and science return of new missions. Most of the high-return missions
possible with current capabilities have been done already or are being done now. Table 1 lists
some of the most challenging current and future missions that DSMS is supporting or may have
to support.
Table 1
CURRENT AND FUTURE MISSION CHALLENGES
Mission
Characteristics and Challenges
Mars Reconnaissance Orbiter
Ka-band downlink, optical navigation experiment, high precision
navigation in orbit
Dawn (minor planet orbiter)
Low-thrust, optical navigation
Phoenix (Mars polar lander)
Unbalanced ACS thrusters, difficult communication geometry
Mars Science Laboratory (rover)
Requirement to be able to land at a high elevation, EDL with a
heavier landing vehicle, hazard avoidance
Juno (Jupiter polar orbiter)
Ka-band tracking
Mars Scout – landers/rovers
Increased reliance on in-situ navigation means, pinpoint landing,
hazard avoidance
Mars com/nav relay orbiter
UHF, X-, Ka-band and/or optical links
Mars Sample Return
Pinpoint landing, Mars ascent, Mars-orbit rendezvous
Mars Scout – aero-rovers
Atmospheric navigation at Mars
Comet surface sampling mission
Close-proximity operations in an unpredictable environment, flight
path/attitude control interaction, pinpoint landing
Outer-planet moon orbiter
Three-body navigation, radiation environment, long round-trip light
times
Multiple-spacecraft telescopes
Precision formation flying
Venus in-situ explorers
EDL and/or atmospheric navigation at Venus
Mars human precursor missions
Demonstration of highly reliable capabilities
Outer-planet moon lander
Three-body navigation and EDL, radiation environment
2
The main trends and challenges that can be derived from these missions are:
1. Increased challenge from round-trip light time, either because operations are done further
away from the Earth, or because the kind of things that can be done are limited if the
ground has to be in the loop.
2. Increased use of in-situ and autonomous navigation: Mars in-situ relays, optical
navigation, close-proximity operations.
3. Increased use of low-thrust propulsion and low-energy transfers using three- or four-body
dynamics.
4. Increased need for higher accuracy in guidance, navigation, and control.
5. Increased need for integration between flight path and attitude control.
In addition there is always the desire to decrease the cost and risk of high-performance
navigation, by producing low-weight, low-cost, highly reliable navigation components. The
availability of these components will enable missions that would not be possible or affordable
otherwise.
DSMS also performs the same kind of analysis for deep-space communication capabilities3,
and it is proposing improvements in the deep-space communications infrastructure that can also
benefit navigation. Some of the trends in the communications side that can be leveraged by
navigation users are:
1. Use of arrays of many smaller radio antennas.
2. Transition to higher frequency bands, in particular Ka-band and possibly optical.
3. Increased reliance on in-situ communication links.
4. Increased inter-operability with other science and space agencies.
ENABLING STRATEGIES
Many of the navigation capabilities required by future missions are currently available, but
some new missions have requirements that cannot be fulfilled without improving existing
capabilities or developing new technologies. The following are the strategies that DSMS plans to
follow in order to improve deep-space navigation capabilities.
Advance Radio-Metric Tracking Capabilities
DSMS will leverage the improvements in communication capabilities in order to benefit its
navigation users, including the migration to the Ka frequency band and the deployment of large
arrays of smaller antennas.
Arrays of smaller antennas can benefit navigation in multiple ways. A large array can have
more sensitivity than a single antenna, and allow for uninterrupted tracking during many missioncritical phases, including entry, descent and landing (EDL). The array size can be adjusted o
fulfill the particular requirements of a mission for each of its mission phases, so fewer antennas
can be used when the only requirement is for range and Doppler, and more when high-rate
telemetry is also needed. If the number of antennas is high enough, it could be possible to have
continuous Doppler-only tracking for many spacecraft, without tying up a bigger antenna, and
also to have continuous tracking of radio-frequency sources for fast Earth orientation calibration.
Smaller antennas have a wider beam, so multiple close spacecraft (e.g. all the spacecraft at or
3
approaching Mars) and natural radio-frequency sources can be tracked simultaneously and new
tracking techniques, such as Same-Beam Interferometry, can be used routinely. In addition
weaker and therefore more numerous and closer quasar sources could be used with the increased
sensitivity, increasing the Very Long Baseline Interferometry (VLBI) measurement accuracy.
Antenna arrays can also be used to precisely determine the angular position of spacecraft near the
Earth and are particularly useful for early orbit search and acquisition. The increased
communication rates can be used to rapidly download higher volumes of more detailed optical
navigation pictures and spacecraft small-force and attitude data.
Ka-band tracking has the advantage that it is less sensitive to charged-particle effects and has
a smaller measurement noise. Moreover, radio-frequency sources are more compact in Ka-band.
However, in order to be able to use Ka-band tracking at its full potential it will be necessary to
improve the accuracy and timeliness of neutral media and Earth orientation calibrations, and to
obtain a more precise quasar catalogue for Ka-band sources.
Sharing antenna assets with other science and space agencies will increase the availability of
spacecraft VLBI tracking. Today, using only DSN assets, it is only possible to perform VLBI
tracking of a given spacecraft twice a day in two somewhat narrow temporal windows when two
DSN ground complexes have common visibility of the spacecraft. Using additional sites we
could obtain improved observation geometry, when using distant antennas with the same
approximate longitude, and have many additional tracking opportunities.
In the future the DSN should be able to perform pseudo-noise ranging, in order to increase
ranging accuracy and reduce the impact of ranging on data rates; and it should improve the
accuracy of range calibrations, by performing automated, multiple-frequency antenna delay
calibrations. Increased DSN automation, required to efficiently and affordably operate large
arrays of antennas, would also benefit navigation by allowing navigation users to reliably
optimize the tracking parameters in order to accommodate changes in the communication link
and the accuracy needs.
For the space segment we should pursue the development of light-weight multi-band digital
transponders that can regenerate the range signal, and can use pseudo-noise range coding.
Expand the Use of Optical Navigation
Optical navigation (opnav) is the use of an onboard camera to image solar system bodies
against a star background in order to improve the knowledge of the spacecraft’s angular position
relative to that body. The solar system body can be a planet, planetary satellites, or minor planets
such as asteroids or comets. This data type can be used on its own, or as is more common,
combined with radiometric data types to compute a navigation solution.
The data type is most useful when the ephemeris of the target body is not known to high
accuracy. Thus, it has been extensively used for missions to the outer planets, either for flybys as
for Voyager at Jupiter, Saturn, Uranus and Neptune, and Galileo and Cassini for satellite tours.
For Voyager, opnav was used to image the large satellites in order to pinpoint the location of the
spacecraft relative to the planet (a technique which can also be applied at Mars, as has been
demonstrated by Viking and will be again for the Mars Reconnaissance Orbiter during its
approach to Mars). For Galileo and Cassini, opnav uses image of the satellites in order to
accurately target its flybys of those bodies, as well as help improve the overall ephemeris
accuracies of the satellites.
4
Opnav is especially critical for missions to small bodies because their ground based
ephemerides are only accurate to the tens to hundreds of km for asteroids, and up thousands for
comets. The Galileo encounters of Gaspra and Ida, NEAR, Deep Space 1 (DS1), Stardust, and
Deep Impact (DI) missions all could not have been possible without it. In addition to the flybys
of these bodies, opnav is necessary for characterizing the shape, orientation, and spin rate of these
bodies, which is critical for close approaches or landings. Optical images are also used for
planetary Entry Descent, and Landing (EDL), such as the DIMES system used on the Mars
Exploration Rovers for reducing the lateral velocity before impact. Future requirements to
perform pinpoint landings on any of these targets drive the continuing need for cameras.
Historically, opnav has been performed using cameras whose primary purpose is as a science
instrument. The trend however, is to develop a dedicated opnav camera which is affordable both
in terms of cost, mass, and power. Such camera is being flown as a demo on MRO. This camera
has a focal length of 500 mm, an aperture of 60 mm, uses 3-5 W of power, but weighs only 2.8
kg. With improvements in ground processing, the accuracy of this should rival that of Cassini’s
25 kg camera. Further improvements in a dedicated camera would be to place it gimbals, which
reduces or eliminates the need to slew the entire spacecraft in order to take images.
Improvements in onboard image processing methods to compress large images and extract out
relevant information without sending back the entire image also can improve the amount of opnav
data sent back, thus contributing to improved navigation results. This kind of camera will make
opnav possible as main or complementary navigation type for many missions, including small
missions that need to navigate autonomously in the proximity of small bodies.
As a related but distinct topic, the DSN is also considering the use of optical frequencies for
communications with deep space assets, and in that case there could also be the possibility of
adding metric capabilities to the communication system in order to measure range very
accurately. The big difference with respect to near-Earth laser ranging systems is that, due to the
great distance, corner cube reflectors could not return enough signal, so a regenerative optical
transponder would need to be used, with the difficulty of reliably calibrating transponder delays
in order to be able to obtain high-precision measurements.
Re-engineer the Navigation Toolset
Most of the software tools that are currently being used operationally at JPL for navigation
and mission design have evolved from tools that were first developed in the late 1960s and 1970s,
so they still have the limitations associated with the FORTRAN/mainframe software development
paradigm. At that time memory, disk space, processing resources, and bandwidth were scarce, so
the software and the interfaces were optimized to be economical with all of those resources, at the
expense of maintainability and usability. Nowadays we have in our laptops better performance in
every measure than that of the best computers in the world thirty years ago. That is why we need
to re-engineer the navigation and trajectory design software set in order to take advantage of
rapidly evolving hardware capabilities, including parallel processing, and to enable further
ground automation. The legacy software is difficult to maintain and extend, and cannot be easily
adapted to take advantage of modern computing paradigms. We need to re-engineer the software
to be able to use modern software engineering approaches that reduce extension and maintenance
costs, and to enable high- and low-level parallelization using multiple processors.
DSMS is moving towards a modern Information Systems Architecture (ISA), based on three
main tenets4: systems are composed of a loosely coupled collection of well-defined and
interoperable network-accessible services and tools; information is defined externally to any
given service or tool and is readily available to all authorized users; client interfaces and displays
5
are decoupled from their underlying processing and data management functions. DSMS is also
participating in the Consultative Committee for Space Data Systems5 (CCSDS), working with
other space agencies to develop recommendations and standards for space systems, in order to
further interoperability in the international space community. As we re-engineer the navigation
and mission design software, we will base it on the DSMS ISA infrastructure paradigm and use as
much as possible standard interfaces defined by the CCSDS.
In addition we will need to add new modeling, processing, analysis, and optimization
capabilities that can enable new types of deep-space missions, including low-thrust and lowenergy trajectory design and control, tools for integrated optimization and control of closely
coupled GNC systems, enhanced analysis capabilities for aero-assist and in-situ architectures, and
multi-mission autonomous navigation tools.
Develop General-Purpose Autonomous Navigation Capabilities
Autonomous Navigation (Autonav) for spacecraft has two primary benefits; one is to
enhance or enable missions which would otherwise not be possible due to round-trip light time or
other limitations, the other is to potentially reduce costs by reducing the number of people needed
for routine navigation operations. Autonav was first used on the DS1 mission as a technology
experiment. DS1 used optical images of asteroids to perform orbit determination and then guide
the spacecraft’s trajectory by performing maneuvers, either with long term control of the ion
engines or with impulsive delta-vs using hydrazine thrusters. Both were successfully
demonstrated during the interplanetary cruise phase of the mission. A subset of this software was
then used to control camera pointing during its flyby of the comet Borrelly; this same software
was also used for the Stardust flyby of comet Wild 2. A heritage DS1 Autonav system was used
by DI impactor to control its trajectory to a high enough accuracy to hit a lit side of Tempel 1, as
well as image the resultant crater from the flyby spacecraft.
The above missions show examples of the mission enhancing capabilities of Autonav. For
example, in all the comet flybys, most or all of the frames have the target in them as compared to
a non-autonomous flyby, such as the Galileo Gaspra or Ida flybys, where dozens of images were
taken in order to capture the target in a handful of frames. For the DI impactor, Autonav was
enabling in that the comet had to be resolved in order to hit a lit spot on the nucleus; this only
occurred several hours before impact which precluded ground control due to the light time. The
DS1 cruise case illustrates an example where Autonav could reduce the number of ground
navigators needed for a during a long and relatively quiescent cruise phase.
Up to now, Autonav use has been limited to small body flyby and impact missions.
However, there are many new missions and mission phases of very high exploration and
scientific value, such as ascent, pinpoint landing, deep-space or Mars-orbit rendezvous, and deepspace formation flying, that require more reliance on close-loop integrated six-degrees-offreedom control of attitude and trajectory, and that cannot be performed with the ground in the
loop. Thus, in order to fully realize the benefits of Autonav, the current DS1/DI heritage system
needs to be re-engineered to make it more general and applicable to a wider range of missions.
For example, the current system only uses optical data, whereas future missions will need
additional data types such as LIDAR, or in situ radiometric data. The integration of new
hardware with the software, such as the Inertial Stellar Compass, which can provide attitude
determination as well as translational information if combined with Autonav, is also being
pursued. The experimental MRO opnav camera is another piece of hardware which can be folded
into an Autonav system; this camera combined with a gimbal could make it very attractive for
6
lower-cost Discovery or Mars Scout missions to autonomously and safely operate in close
proximity to a small body, or to land or navigate at Mars.
Improve Frequency and Timing Systems
Highly-stable frequency standards reduce the need for two-way data types and, because there
is no need to close the communication link, free users from round-trip light time constraints.
They also allow multiple users to be served by just one asset, decreasing the cost of supporting
multiple spacecraft. Currently multiple Mars spacecraft can be tracked simultaneously using just
one DSN antenna; but only one of them can be tracked in two-way mode. Only when the
spacecraft tracked in one-way mode has an ultra-stable oscillator (USO), like Mars Global
Surveyor (MGS) does, is the one-way data suitable for trajectory determination. In the future we
could have in-situ GPS-like systems at the Moon or Mars, with a few high-orbit spacecraft
serving many low-orbit and landed assets, and USOs would be needed at least for the service
provider spacecraft, in order for the users to be able to obtain real-time trajectory and timing
solutions. Furthermore, the user spacecraft should also have USOs in order to reduce the number
of required service-provider spacecraft and to enable the use of sparse data with dynamics-based
orbit determination approaches.
In addition, many parts of the Frequency and Timing Subsystem (FTS) of the DSN are
obsolescent. The systematic upgrading of the FTS on an element-by-element basis, which has
been underway for several years, should be continued.
Develop In-situ Tracking Infrastructure
NASA already has multiple assets at Mars providing – MGS, Odyssey – and using – Mars
Exploration Rovers – in-situ communications and navigation capabilities. In-situ radio-metric
tracking was used by the MER mission in order to improve the position determination of the
landed rovers, and to allow for the imaging of the rovers by MGS. MRO is hosting a proximity
radio that can be used by other future missions. Phoenix and Mars Science Laboratory are
planning to use in-situ radio-metrics for EDL reconstruction and landed position determination.
In the future we may have high-orbit dedicated relay spacecraft that provide longer tracking
passes for landed or low-orbit assets, and that can track spacecraft arriving or departing Mars.
We are developing radios and software that can exploit these in-situ capabilities, and can enable
high-precision trajectory and position determination, even for real-time application, such as
during guided EDL for pinpoint landing.
These kinds of capabilities may also be available at other bodies of interest, starting with the
Moon in order to allow for global communications and navigation capabilities in support of Lunar
exploration, and eventually at other solar system bodies such as Europa or Titan, in order to allow
for simpler landed assets, more precise navigation, and increased data rates.
CONCLUSION
In the future NASA is going to develop and operate missions that require navigation
capabilities that are beyond what is available today. DSMS has identified the main challenges for
these future missions, and developed strategies to enable the successful operation of these
missions. The capabilities enabled by these strategies will also be available to other NASA
centers and applications, such as Human and Robotic Lunar Exploration, so they do not have to
be developed from scratch, and can reduce schedule, risk and cost.
7
ACKNOWLEDGMENT
We would like to express our appreciation for the enthusiastic support from Barry
Geldzahler, the DSMS Program Executive, to the advancement of deep-space communication and
navigation capabilities. We would also like to acknowledge the team, led by Les Deutsch, which
developed the DSMS Roadmap; and the many contributors from the JPL Guidance, Navigation,
and Control Section, the Tracking Systems and Applications Section and other JPL Sections and
Offices for their help developing the Navigation Vision.
This research was carried out at the Jet Propulsion Laboratory, California Institute of
Technology, under a contract with the National Aeronautics and Space Administration.
Reference herein to any specific commercial product, process, or service by trade name,
trademark, manufacturer, or otherwise, does not constitute or imply its endorsement by the United
States Government or the Jet Propulsion Laboratory, California Institute of Technology.
REFERENCES
1. Abraham, D.S.; “DSMS Mission Set Trends.” Deep Space Mission System (DSMS)
Strategic Planning Retreat, November 15-16, 2005.
2. Cesarone, R.J., and Abraham, D.S.; “Long Range Planning for the Deep Space Network”,
AIAA Space 2003 Conference paper AIAA-2003-6416, Long Beach, CA, September 23-25,
2003.
3. Weber, W.J., Cesarone, R.J., Abraham, D.S., Doms, P.E., Doyle, R.J, Edwards, C.D, Hooke,
A.J, Lesh J.R., and Miller, R.B. ; “Transforming the Deep Space Network into the
Interplanetary Network”, International Astronautical Congress paper IAC-03-0.4.01, Bremen,
Germany, September 29 – October 3, 2003.
4. McVittie, T.; “DSMS Software Architecture Overview”, presentation to the 369 Section,
October 20, 2003.
5. Consultative Committee for Space Data Systems, http://www.ccsds.org/
8