Author Topic: Mars GPS system  (Read 17249 times)

Offline savuporo

  • Senior Member
  • *****
  • Posts: 5155
  • Liked: 979
  • Likes Given: 345
Re: Mars GPS system
« Reply #40 on: 05/08/2016 03:56 AM »
Experiments have been done using the current Mars fleet to locate the rovers. I think they used the Doppler effect to get triangulation using a single satellite.

All kinds of cleverness is possible.

http://scpnt.stanford.edu/pnt/PNT11/2011_presentation_files/13_Whalen-PNT2011.pdf
"Boeing Timing & Location"
An Indoor Capable Time Transfer and Geolocation System
Orion - the first and only manned not-too-deep-space craft

Offline ChrisWilson68

  • Senior Member
  • *****
  • Posts: 3558
  • Sunnyvale, CA
  • Liked: 2126
  • Likes Given: 2535
Re: Mars GPS system
« Reply #41 on: 05/08/2016 04:36 AM »
The thing we haven't done is land several spacecraft at/near the same location. 
One-off are considerably easier.

My impression is that the biggest issue for landing precision is not that the craft doesn't know its position but that it doesn't have the control system to make small adjustments necessary on the way down to hit a precise spot.  An unmanned spacecraft aerobraking behind a heat shield or coming down under a parachute is typically completely passive and unable to modify its course.

That will probably change with Red Dragon.

We already have the entire Red Planet mapped from orbit with such a high degree of precision that a modern inertial navigation system combined with some cheap cameras and some really good software are probably all that are needed to put a lander right next to a previous lander.  No GPS necessary.

Rovers will also be able to tell very precisely where they are based on just looking at the terrain and comparing that to the mapping data we already have.

Offline savuporo

  • Senior Member
  • *****
  • Posts: 5155
  • Liked: 979
  • Likes Given: 345
Re: Mars GPS system
« Reply #42 on: 05/08/2016 04:43 AM »
My impression is that the biggest issue for landing precision is not that the craft doesn't know its position but that it doesn't have the control system to make small adjustments necessary on the way down to hit a precise spot. ..
Actually, localization is the bigger issue.
Orion - the first and only manned not-too-deep-space craft

Offline CuddlyRocket

Re: Mars GPS system
« Reply #43 on: 05/08/2016 05:45 PM »
Since at least SpaceX needs to set up a high density communication network anyway, that network could double as positioning system.

A GPS system requires more than just the satellites; there's quite a bit of ground-based equipment and personnel involved in keeping it running.

However, having thought about this a bit, I realised that the Earth-based equipment and personnel might be just that even for a Martian GPS (you'd possibly need some equipment on Mars)! I don't know if low latency between ground and satellites is a necessary requirement.

Offline Space Ghost 1962

  • Senior Member
  • *****
  • Posts: 2674
  • Whatcha gonna do when the Ghost zaps you?
  • Liked: 2667
  • Likes Given: 2100
Re: Mars GPS system
« Reply #44 on: 05/08/2016 06:17 PM »
Keep in mind Mars is a very different environment then Earth, especially for radio comm.

To begin with little radiative/EMI noise, and most of that coming from the plasma wake, which is much closer than Earth's, plus no Van Allen belts. We already have a significant reflection from the magnetosphere a few thousand kilometers with the boundary below it.

Your ground stations will be in your equipment, and they will be clustered in locations (e.g. a steerable dish on SC would allow low power multichannel two way link mobile devices leveraging a high powered sat).

We've got no propagation studies - if we did we might find that MF/LF might be more effective, getting around LOS issues for most surface needs, thus a comm/nav architecture that leverages existing on orbit assets for existing exploration landers/orbiters/rovers is "good enough" back channel for the foreseeable future.

Offline A_M_Swallow

  • Elite Veteran
  • Senior Member
  • *****
  • Posts: 8224
  • South coast of England
  • Liked: 268
  • Likes Given: 108
Re: Mars GPS system
« Reply #45 on: 05/08/2016 10:49 PM »
Can a single ground station provide the reference signal to keep the satellites synchronised?

We may need to know the location of this station with extreme accuracy, possibly by defining as a base point.

Offline savuporo

  • Senior Member
  • *****
  • Posts: 5155
  • Liked: 979
  • Likes Given: 345
Re: Mars GPS system
« Reply #46 on: 05/09/2016 12:43 AM »
Ground beacon based references have been studied. The key question is if you can get radio signal during the most critical time, from the perspective of error accumulation - i.e. when the vehicle is flying hypersonic lift in the plasma sheath. For MSL, this was a 'closed loop' flight phase where bank angle is modulated against internal position estimate that is based on IMU only. It's 'closed loop' relative to internal model, not against some absolute position measurement.
This is the phase of flight where most of the downrange positional estimate error accumulates, from the hand off point from ground based tracking.
It does appear that UHF radio could be used and at least in theory, it could be used to practically nullify the largest uncertainty variables ( mostly caused by atmospheric density and pressure variations ) from the equation

For example : https://arxiv.org/ftp/arxiv/papers/1507/1507.00937.pdf

« Last Edit: 05/09/2016 12:46 AM by savuporo »
Orion - the first and only manned not-too-deep-space craft

Offline virnin

  • Member
  • Posts: 95
  • Kansas
  • Liked: 29
  • Likes Given: 38
Re: Mars GPS system
« Reply #47 on: 05/10/2016 01:28 AM »
The tiny atomic clocks are attractive but how much will you spend (power/mass) keeping them within their operating temperature range aboard the satellite?   "operates from -10 C to +35 C"

Offline savuporo

  • Senior Member
  • *****
  • Posts: 5155
  • Liked: 979
  • Likes Given: 345
Re: Mars GPS system
« Reply #48 on: 05/10/2016 03:23 AM »
The tiny atomic clocks are attractive but how much will you spend (power/mass) keeping them within their operating temperature range aboard the satellite?   "operates from -10 C to +35 C"

You spend a few budget cycles maturing the tech and trying to get this flight demoed

http://www.nasa.gov/mission_pages/tdm/clock/index.html
Orion - the first and only manned not-too-deep-space craft

Offline sanman

  • Senior Member
  • *****
  • Posts: 3998
  • Liked: 550
  • Likes Given: 7
Re: Mars GPS system
« Reply #49 on: 05/11/2016 08:28 AM »
Came across an article about achieving precision navigation without GPS:

http://www.draper.com/news/navigation-even-without-gps

So it may not be necessary to have a GPS system in place to get comparable capabilities.

Offline Hotblack Desiato

  • Full Member
  • ***
  • Posts: 345
  • Austria
  • Liked: 64
  • Likes Given: 49
Re: Mars GPS system
« Reply #50 on: 05/11/2016 09:06 AM »
Regarding ground stations and so, and how accurat a gps with only one or two know ground locations is.

If there are enough satellites in orbit, they could measure their relative position to each other. That would give them accurate orbits within a few meters (they could even measure the influence of phobos, which might be a problem sooner or later).

They would require just two (actually 3, if we don't even know in which direction the sats rotate) positions on the surface of Mars (should be close enough to be seen simultanously), and the whole network knows the exact position of every sat relative to the surface.

And then, there is a spot on Mars called Airy-0. That's basically the martian greenwich point. So even if those surface stations are not exactly positioned, we just define them as being exact positioned (It would be a bit more interesting to just place a surface station or rover at Airy-0, maybe even drive at the exact spot, and it's done).

More important later will be the relative position to a point of interest. For example a manned station, maybe even a colony: it's nice to know where the station is, but for rovers etc in close proximity to that station, it's more important to know the relative position to the station or facilities inside that station than to know where exactly (1-10m accuracy) the station and/or the rover is on mars. So even a relatively non-accurat positioning of 100-500m accuracy could be sufficient.

Besides, let's assume, SpaceX proceeds with the plan of colonizing Mars and they set up an internet relay sat constellation, they could send 3-5 sats with every Red Dragon (There is still a trunk on that craft), after 10 RDs, they are done with that constellation. And a positioning service could be added to those sats without much extra work.

(Just a fancy idea: that could be done on Earth too, resulting in the first privately owned GPS-system, and with their proposed 4000 sat constellation, at every moment, receivers would see at least a hundred sats, with interesting effects on the accuracy)

Offline Robotbeat

  • Senior Member
  • *****
  • Posts: 27332
  • Minnesota
  • Liked: 7275
  • Likes Given: 4990
Re: Mars GPS system
« Reply #51 on: 05/12/2016 12:49 AM »
Ground beacon based references have been studied. The key question is if you can get radio signal during the most critical time, from the perspective of error accumulation - i.e. when the vehicle is flying hypersonic lift in the plasma sheath. For MSL, this was a 'closed loop' flight phase where bank angle is modulated against internal position estimate that is based on IMU only. It's 'closed loop' relative to internal model, not against some absolute position measurement.
This is the phase of flight where most of the downrange positional estimate error accumulates, from the hand off point from ground based tracking.
It does appear that UHF radio could be used and at least in theory, it could be used to practically nullify the largest uncertainty variables ( mostly caused by atmospheric density and pressure variations ) from the equation

For example : https://arxiv.org/ftp/arxiv/papers/1507/1507.00937.pdf
Hypersonic reentry and plasma blackout doesn't last very long. Dead reckoning can be used, and if sensitive enough you would only drift a few tens of meters or so (laser ring gyros might bring this down to a few meters). This could be nulled out aerodynamically pretty quick after the capsule drops out of plasma blackout (lasts like a minute?).

You wouldn't want to rely on dead reckoning the whole way, that's for sure, but if it's JUST during plasma blackout, you should be fine.
Chris  Whoever loves correction loves knowledge, but he who hates reproof is stupid.

To the maximum extent practicable, the Federal Government shall plan missions to accommodate the space transportation services capabilities of United States commercial providers. US law http://goo.gl/YZYNt0

Offline savuporo

  • Senior Member
  • *****
  • Posts: 5155
  • Liked: 979
  • Likes Given: 345
Re: Mars GPS system
« Reply #52 on: 05/12/2016 04:19 AM »
Ground beacon based references have been studied. The key question is if you can get radio signal during the most critical time, from the perspective of error accumulation - i.e. when the vehicle is flying hypersonic lift in the plasma sheath. For MSL, this was a 'closed loop' flight phase where bank angle is modulated against internal position estimate that is based on IMU only. It's 'closed loop' relative to internal model, not against some absolute position measurement.
This is the phase of flight where most of the downrange positional estimate error accumulates, from the hand off point from ground based tracking.
It does appear that UHF radio could be used and at least in theory, it could be used to practically nullify the largest uncertainty variables ( mostly caused by atmospheric density and pressure variations ) from the equation

For example : https://arxiv.org/ftp/arxiv/papers/1507/1507.00937.pdf
Hypersonic reentry and plasma blackout doesn't last very long. Dead reckoning can be used, and if sensitive enough you would only drift a few tens of meters or so (laser ring gyros might bring this down to a few meters). This could be nulled out aerodynamically pretty quick after the capsule drops out of plasma blackout (lasts like a minute?).

You wouldn't want to rely on dead reckoning the whole way, that's for sure, but if it's JUST during plasma blackout, you should be fine.
IMU based navigation is not dead reckoning. Dead reckoning in GN&C and robotics means estimating your pose based on elapsed time on control signals without feedback of any sort.
IMU navigation is still closed loop, with a well quantifiable uncertainty and error propagation characteristics.
And no, with FOGs a few tens of meters of positional error over the  hypersonic phase is not really doable, you an get close to tens of meters/s in velocity. RLGs (combined with FOGs) would help, but not that much.
Orion - the first and only manned not-too-deep-space craft

Offline Robotbeat

  • Senior Member
  • *****
  • Posts: 27332
  • Minnesota
  • Liked: 7275
  • Likes Given: 4990
Re: Mars GPS system
« Reply #53 on: 05/12/2016 01:50 PM »
Ground beacon based references have been studied. The key question is if you can get radio signal during the most critical time, from the perspective of error accumulation - i.e. when the vehicle is flying hypersonic lift in the plasma sheath. For MSL, this was a 'closed loop' flight phase where bank angle is modulated against internal position estimate that is based on IMU only. It's 'closed loop' relative to internal model, not against some absolute position measurement.
This is the phase of flight where most of the downrange positional estimate error accumulates, from the hand off point from ground based tracking.
It does appear that UHF radio could be used and at least in theory, it could be used to practically nullify the largest uncertainty variables ( mostly caused by atmospheric density and pressure variations ) from the equation

For example : https://arxiv.org/ftp/arxiv/papers/1507/1507.00937.pdf
Hypersonic reentry and plasma blackout doesn't last very long. Dead reckoning can be used, and if sensitive enough you would only drift a few tens of meters or so (laser ring gyros might bring this down to a few meters). This could be nulled out aerodynamically pretty quick after the capsule drops out of plasma blackout (lasts like a minute?).

You wouldn't want to rely on dead reckoning the whole way, that's for sure, but if it's JUST during plasma blackout, you should be fine.
IMU based navigation is not dead reckoning. Dead reckoning in GN&C and robotics means estimating your pose based on elapsed time on control signals without feedback of any sort.
IMU navigation is still closed loop, with a well quantifiable uncertainty and error propagation characteristics.
And no, with FOGs a few tens of meters of positional error over the  hypersonic phase is not really doable, you an get close to tens of meters/s in velocity. RLGs (combined with FOGs) would help, but not that much.
Thanks for the clarification on the proper use of the term "dead reckoning."
Chris  Whoever loves correction loves knowledge, but he who hates reproof is stupid.

To the maximum extent practicable, the Federal Government shall plan missions to accommodate the space transportation services capabilities of United States commercial providers. US law http://goo.gl/YZYNt0

Offline Robotbeat

  • Senior Member
  • *****
  • Posts: 27332
  • Minnesota
  • Liked: 7275
  • Likes Given: 4990
Re: Mars GPS system
« Reply #54 on: 05/12/2016 06:58 PM »
Plasma wake is only generated at certain speeds (and to some degree, air pressure). Has to be high hypersonic. Not an issue when merely supersonic.
Chris  Whoever loves correction loves knowledge, but he who hates reproof is stupid.

To the maximum extent practicable, the Federal Government shall plan missions to accommodate the space transportation services capabilities of United States commercial providers. US law http://goo.gl/YZYNt0

Offline redliox

  • Full Member
  • ****
  • Posts: 1777
  • Arizona USA
  • Liked: 332
  • Likes Given: 56
Re: Mars GPS system
« Reply #55 on: 05/13/2016 09:57 PM »
cuddihy made an interesting point regarding orbits at Mars, especially stationary orbits.  His points also came from the Planetary Society's Emily Lakdawalla.  Here's cuddihy's stuff:

Yes, mars geostationary orbit "a" is 20,438 km, but then that can only cover abt 1/3 of the surface (like a GSO sat). If you're willing to launch 3 or 4 of them that works, but if you go to about 8500-9000 km you would get several hours coverage over a landing site per pass from a single satellite, with global mars coverage but only a few hours per site at a time.

Either one of these altitudes gets nearly full earth coverage, and with big enough antennas like they can build now you could get impressive, multi - MB coverage to earth continuously.

So in the near term the medium orbit single sat is probably the best idea, because it alone can give you hours long high bw coverage for hours for important things like EDL.
 Long term, it makes sense to put 3 MSO sats around mars.

The problem is you can't launch any of that in the same launch as red dragon 2018--it's going to get there at the same time as red dragon and aero braking or manuevering into final orbit/checkout makes it useless for EDL. So for now you're stuck with the low orbit Sats relaying...

Optimally we'd have launched a comm sat last week...

Here's a great NASA discussion of martian comm sat orbits:
http://www.planetary.org/blogs/emily-lakdawalla/2013/stationkeeping-in-mars-orbit.html
"Let the trails lead where they may, I will follow."
-Tigatron

Offline Ludus

  • Full Member
  • ****
  • Posts: 911
  • Liked: 354
  • Likes Given: 179
Re: Mars GPS system
« Reply #56 on: 05/14/2016 09:15 PM »
A simple packet transmitted occasionally (around 1 per second) at the SpaceX CommX data rate of 1Gb could give a timing resolution around .001 micro sec and definitely <.01. The Packet contains a timing  and orbital parameter set. The remaining requirement is high accuracy clock that is regularly updated from a single stationary ground source (The first Red Dragon maybe). This utilizes the high data rate comm system to also perform accurate GPS functionality. This Gb data comm system is a 10-20x data rate speed increase over any current operating sat data comm system. So this was not a previous option.

A +-.01 micro sec error is +- 3m in distance. The older comm data systems would be at best around +-30m in accuracy implemented in the same manner.

Now the final item is low packet jitter handling in the transmitter and receiver modems. This will probably be the highest source of timing error.

So a secondary function of the comm sats with Gb data rates capabilities could be a cyclic GPS data packet transmitted. The sats would then require a high accuracy clock (low drift over ~16 hours of not more than a few micro sec) and the high data rate comm constellation can also perform GPS without having separate receivers or antenna systems. Reducing weight and power usage both on the sats and on the mobile systems.

I think there's very nearly zero chance that any human gets anywhere near Mars without without an IP/GPS/observation satellite constellation of this sort in place first.

Online AncientU

  • Senior Member
  • *****
  • Posts: 5470
  • Liked: 3319
  • Likes Given: 4703
Re: Mars GPS system
« Reply #57 on: 05/15/2016 12:38 PM »
A simple packet transmitted occasionally (around 1 per second) at the SpaceX CommX data rate of 1Gb could give a timing resolution around .001 micro sec and definitely <.01. The Packet contains a timing  and orbital parameter set. The remaining requirement is high accuracy clock that is regularly updated from a single stationary ground source (The first Red Dragon maybe). This utilizes the high data rate comm system to also perform accurate GPS functionality. This Gb data comm system is a 10-20x data rate speed increase over any current operating sat data comm system. So this was not a previous option.

A +-.01 micro sec error is +- 3m in distance. The older comm data systems would be at best around +-30m in accuracy implemented in the same manner.

Now the final item is low packet jitter handling in the transmitter and receiver modems. This will probably be the highest source of timing error.

So a secondary function of the comm sats with Gb data rates capabilities could be a cyclic GPS data packet transmitted. The sats would then require a high accuracy clock (low drift over ~16 hours of not more than a few micro sec) and the high data rate comm constellation can also perform GPS without having separate receivers or antenna systems. Reducing weight and power usage both on the sats and on the mobile systems.

I think there's very nearly zero chance that any human gets anywhere near Mars without without an IP/GPS/observation satellite constellation of this sort in place first.

I agree, but think that both a GPS and high bandwidth comms (same satellites with dual purpose) will be available for the build-up of capabilities before humans arrive.  Robotics, fully automated ISRU systems, and sophisticated rovers will need orders of magnitude higher bandwidth and certainty of location from the start to coordinate their efforts.  Using the Red Dragons for GPS ground stations is a perfect payload for these missions.
"If we shared everything [we are working on] people would think we are insane!"
-- SpaceX friend of mlindner

Offline oldAtlas_Eguy

  • Senior Member
  • *****
  • Posts: 3291
  • Florida
  • Liked: 1692
  • Likes Given: 192
Re: Mars GPS system
« Reply #58 on: 05/15/2016 06:39 PM »
A simple packet transmitted occasionally (around 1 per second) at the SpaceX CommX data rate of 1Gb could give a timing resolution around .001 micro sec and definitely <.01. The Packet contains a timing  and orbital parameter set. The remaining requirement is high accuracy clock that is regularly updated from a single stationary ground source (The first Red Dragon maybe). This utilizes the high data rate comm system to also perform accurate GPS functionality. This Gb data comm system is a 10-20x data rate speed increase over any current operating sat data comm system. So this was not a previous option.

A +-.01 micro sec error is +- 3m in distance. The older comm data systems would be at best around +-30m in accuracy implemented in the same manner.

Now the final item is low packet jitter handling in the transmitter and receiver modems. This will probably be the highest source of timing error.

So a secondary function of the comm sats with Gb data rates capabilities could be a cyclic GPS data packet transmitted. The sats would then require a high accuracy clock (low drift over ~16 hours of not more than a few micro sec) and the high data rate comm constellation can also perform GPS without having separate receivers or antenna systems. Reducing weight and power usage both on the sats and on the mobile systems.

I think there's very nearly zero chance that any human gets anywhere near Mars without without an IP/GPS/observation satellite constellation of this sort in place first.

I agree, but think that both a GPS and high bandwidth comms (same satellites with dual purpose) will be available for the build-up of capabilities before humans arrive.  Robotics, fully automated ISRU systems, and sophisticated rovers will need orders of magnitude higher bandwidth and certainty of location from the start to coordinate their efforts.  Using the Red Dragons for GPS ground stations is a perfect payload for these missions.
If you think of GPS as nothing more than a supper accurate Internet timing service then the need for a separate set of systems for GPS disappears. It then becomes just one more standard Internet Data service provided by a high speed server on each comm sat. It would only take about 16 to 32 of LMO sats which could be as small as ~50kg at an altitude of ~500km. Add a pair of larger Earth Link comm sats that then link to one or more of the lower sats and you get both a high accurate full coverage GPS but also a 2Gbs continuous comm link to any one single spot on Mars surface or even in LMO using the comm sats sat-to-sat link capabilities. Total weight budget 1500kg for LMO comm sats and another 1500kg for the two Earth Link sats. 2 F9 launches would do it. Sats would use electric propulsion to get to final orbits after the bus does a Mars orbit capture chemical braking burn. On orbit life would be > than 2 synods at which time more capable and bigger sats would replace them. Total costs (no profit) including launch to SpaceX would be ~$80M for 2 F9 launches and another $80 for all the sats which are primarily based on and manufactured using similar parts and subsystems as the Earth comm sats.
NOTE: Data throughput per sat vs the Earth ones would be only 2 Gbps vs Earth ones which could be as high as 16-64Gbps. Lower throughput reduces the overall weight significantly, less power, less heat, less batteries, smaller weight less prop.

The remaining item is that the DSN would not be able to handle these data rates so a new in-orbit system at Earth would be needed to be the Earth port that connects into a high data rate global comm. So at this point DSN would be used for the travel between planets and for probes farther out.

Trying to design a separate GPS system is not a cost effective or mass effective option for Mars. GPS will be a best a hosted system of other critical mission sats like high data rate comms. Although interim software solutions using available assets may be used prior to deployment of these new assets.

Offline Robotbeat

  • Senior Member
  • *****
  • Posts: 27332
  • Minnesota
  • Liked: 7275
  • Likes Given: 4990
Re: Mars GPS system
« Reply #59 on: 11/12/2016 01:41 AM »
Here's a small regional GPS system which would be good for an initial Mars settlement:

https://en.wikipedia.org/wiki/Indian_Regional_Navigation_Satellite_System

It's 7 satellites in GSO and covers India and all around it to 10-20m accuracy (0.1m accuracy encrypted for military use). Probably can get by with 4, especially with a good local clock (which are small and cheap nowadays).

You can actually do just 1 if you use doppler shift.

Ground station could be pretty simple at first.
« Last Edit: 11/12/2016 01:42 AM by Robotbeat »
Chris  Whoever loves correction loves knowledge, but he who hates reproof is stupid.

To the maximum extent practicable, the Federal Government shall plan missions to accommodate the space transportation services capabilities of United States commercial providers. US law http://goo.gl/YZYNt0

Tags: