Starlink: when can we expect Martian deployment?

[edzieba]

Mind you, a GPS satellite at its core is a atom clock (those can be made remarkably small nowadays) that is broadcasting its local time and position. That would not necessarily constitute a trivial modification to Starlink satellites (on top of the bigger solar PV panels), of course.

The clock is only half the solution. The other half is the accurate and up to date measurement of the true orbital parameters of each satellite from a fixed reference (for earth-bound GNSS, that's a reference to fixed ground stations). Orbits will drift and timings will drift - even for atomic clocks - over time, so the system requires constant measurement of those parameters and updating of the satellites that broadcast that up to date orbit reference data along with the timecodes - for Navstar GPS, that would be the Operational Control Segment.

[DanClemmensen]

Mind you, a GPS satellite at its core is a atom clock (those can be made remarkably small nowadays) that is broadcasting its local time and position. That would not necessarily constitute a trivial modification to Starlink satellites (on top of the bigger solar PV panels), of course.

The clock is only half the solution. The other half is the accurate and up to date measurement of the true orbital parameters of each satellite from a fixed reference (for earth-bound GNSS, that's a reference to fixed ground stations). Orbits will drift and timings will drift - even for atomic clocks - over time, so the system requires constant measurement of those parameters and updating of the satellites that broadcast that up to date orbit reference data along with the timecodes - for Navstar GPS, that would be the Operational Control Segment.

GPS is an incredible achievement, but its basic architecture is 45 years old. Adding explicit positioning capability to Starlink could take advantage of existing Starlink features like the ISL links and much more capable onboard computers. Adding hardware-assisted timestamps (conceptually similar to IEEE 1588 but with sub-nanosecond precision) to the ISL links would increase the satellite position accuracy and clock accuracy for all the satellites, averaging out the uncertainties in the positions as determined by the ground control links.

I am not a GPS professional. Basics for those of us who are not GPS professionals: GPS works by allowing the receiver to solve an equation in four unknowns (X,Y, Z, time) given four inputs: the signals from four satellites. Each signal includes the (X,Y,Z,T) of the satellite at the time the signal was transmitted. The receiver does not explicitly know the actual distance to each satellite. Instead, it computes the difference in distances to each satellite by comparing the differences in received time. (With more satellites, you get to solve a system with 4 unknowns and five or more equations. This lets you reduce the uncertainties.) This means that satellite position accuracy is as important as satellite clock accuracy: one microsecond of clock accuracy is equivalent to 300 meters of position accuracy.

[waveney]

Additionally by measuring the detailed errors and corrections needed, the gravity maps of Mars could be very significantly enhanced.  These are very good for detecting hidden features such as buried craters and buried volcanoes.

They do this already on the orbits of MRO (and others) but this would yield much higher detail.

[jak Kennedy]

Whether or not Starlink in it's current configuration will be used in Mars orbit, my question is could they be deployed by Starship after a deorbit burn (if it uses one) even if then placed into a highly elliptical orbit?

[waveney]

Whether or not Starlink in it's current configuration will be used in Mars orbit, my question is could they be deployed by Starship after a deorbit burn (if it uses one) even if then placed into a highly elliptical orbit?

Yes

[Brigantine]

could they be deployed by Starship after a deorbit burn (if it uses one) even if then placed into a highly elliptical orbit?

AIUI the plan was direct EDL from a hyperbolic transfer orbit. So it sounds like your suggestion is to propulsively capture (0.67 km/s) before closest approach, then deploy, then adjust closest approach to enter Mars' atmosphere for EDL.

That might be possible, but at a cost of 20+ tons of payload. (roughly)
You'd need to think about extra propellant storage/conditioning for the propulsive capture, if typical Mars ships don't use the main tanks for the Mars landing burn.

An alternative is to direct "E" from a hyperbolic transfer orbit, Starlinks still on board, but save the "DL" for later.
- Come in from heliocentric orbit at the appropriate Starlink inclination,
- in 1 pass aerocapture/aerobrake to some Mid-Mars-Transfer orbit,
- then at Mid-Mars height, small burn/deploy/small burn and
- then EDL (perhaps in 2 further passes if your destination isn't at a convenient latitude).

This will cost you hours to sols of time, but shouldn't cost much in terms of payload - you may even get a small (<25%) discount on your Starlink payload.

Does that seem right?

[edzieba]

Mind you, a GPS satellite at its core is a atom clock (those can be made remarkably small nowadays) that is broadcasting its local time and position. That would not necessarily constitute a trivial modification to Starlink satellites (on top of the bigger solar PV panels), of course.

The clock is only half the solution. The other half is the accurate and up to date measurement of the true orbital parameters of each satellite from a fixed reference (for earth-bound GNSS, that's a reference to fixed ground stations). Orbits will drift and timings will drift - even for atomic clocks - over time, so the system requires constant measurement of those parameters and updating of the satellites that broadcast that up to date orbit reference data along with the timecodes - for Navstar GPS, that would be the Operational Control Segment.

GPS is an incredible achievement, but its basic architecture is 45 years old. Adding explicit positioning capability to Starlink could take advantage of existing Starlink features like the ISL links and much more capable onboard computers. Adding hardware-assisted timestamps (conceptually similar to IEEE 1588 but with sub-nanosecond precision) to the ISL links would increase the satellite position accuracy and clock accuracy for all the satellites, averaging out the uncertainties in the positions as determined by the ground control links.

That doesn't actually solve the problem: Mars has no groundside measurement reference and control system, and Mars has no independent baseline geodetic survey to align to or from.

Lets assume for a moment that our satellites have magical super-atomic clocks that never ever drift and can ignore relativistic effects, such that we do not need to concern ourselves with satellite-to-satellite relative drift, we can synchronise all the clocks at launch and then ignore sync forever. Lets also assume that the satellites themselves have such exquisitely precise ISL steering gimbals (well above the requirements for laser communicaitons) and truly outstanding laser pulse timing (for accurate ranging) that each satellite can create an exact model of satellite relative positions independently of any ground stations. (Both of these are extremely generous assumptions that either vastly inflate satellite cost or are plain not physically possible).
Even in that situation, all your GNSS system will tell you is your position relative to that satellite constellation. Not your position relative to Mars, which is what you actually care about - as a gross example, a difference in "what time is it" of 1 minute (a difference of what time the satellite constellation thought Mars had when it was set up vs. the actual local time, not a timing error in the constellation itself) would leave you over 14km away if you were standing at Mars' equator. To solve that, you need to accurately align your constellation to the moving surface of Mars. That needs a minimum of 3 independent ground stations, sufficiently spaced to allow trilateration of sufficient accuracy, and those ground stations need a precisely known relationship to each other (e.g. ground-based surveying or some independent triangulation method). Coordinate system alignment is no simple task at planetary scales where assumptions like "gravity always points straight down" or "the ground does not change length" are not strictly true.
If you then incorporate more realistic assumptions, where clocks drift, satellite positions need to tracked at multiple ground sites to allow accurate triangulation, and satellite orbits continuously shift from surface gravity influence and influence from Phobos and Deimos, the demands on the ground segment only increase.

On Earth, its easy to overlook these baseline requirement because we were producing accurate geographic and geodetic surveys centuries before the first satellite was ever launched. On Mars, we do not have that baseline available. We have satellite imagery of the surface, but that is very much not the same as geodetic survey.


A GNSS system for Mars is obviously not impossible, but it is much more complex than just throwing some clocks on some satellites.

[steveleach]

That doesn't actually solve the problem: Mars has no groundside measurement reference and control system, and Mars has no independent baseline geodetic survey to align to or from.

Lets assume for a moment that our satellites have magical super-atomic clocks that never ever drift and can ignore relativistic effects, such that we do not need to concern ourselves with satellite-to-satellite relative drift, we can synchronise all the clocks at launch and then ignore sync forever. Lets also assume that the satellites themselves have such exquisitely precise ISL steering gimbals (well above the requirements for laser communicaitons) and truly outstanding laser pulse timing (for accurate ranging) that each satellite can create an exact model of satellite relative positions independently of any ground stations. (Both of these are extremely generous assumptions that either vastly inflate satellite cost or are plain not physically possible).
Even in that situation, all your GNSS system will tell you is your position relative to that satellite constellation. Not your position relative to Mars, which is what you actually care about - as a gross example, a difference in "what time is it" of 1 minute (a difference of what time the satellite constellation thought Mars had when it was set up vs. the actual local time, not a timing error in the constellation itself) would leave you over 14km away if you were standing at Mars' equator. To solve that, you need to accurately align your constellation to the moving surface of Mars. That needs a minimum of 3 independent ground stations, sufficiently spaced to allow trilateration of sufficient accuracy, and those ground stations need a precisely known relationship to each other (e.g. ground-based surveying or some independent triangulation method). Coordinate system alignment is no simple task at planetary scales where assumptions like "gravity always points straight down" or "the ground does not change length" are not strictly true.
If you then incorporate more realistic assumptions, where clocks drift, satellite positions need to tracked at multiple ground sites to allow accurate triangulation, and satellite orbits continuously shift from surface gravity influence and influence from Phobos and Deimos, the demands on the ground segment only increase.

On Earth, its easy to overlook these baseline requirement because we were producing accurate geographic and geodetic surveys centuries before the first satellite was ever launched. On Mars, we do not have that baseline available. We have satellite imagery of the surface, but that is very much not the same as geodetic survey.


A GNSS system for Mars is obviously not impossible, but it is much more complex than just throwing some clocks on some satellites.

Surely if all the satellites know where they are in relation to each other in X,Y,Z,T (which, as you say, is not trivial) then all you need is a single reference point on the surface that 4 or more sats can reach.

From there, any other point that can be reached by 4 or more sats can be located relative to the reference point, right?

Or am I missing something?

[waveney]

That doesn't actually solve the problem: Mars has no groundside measurement reference and control system, and Mars has no independent baseline geodetic survey to align to or from.

Lets assume for a moment that our satellites have magical super-atomic clocks that never ever drift and can ignore relativistic effects, such that we do not need to concern ourselves with satellite-to-satellite relative drift, we can synchronise all the clocks at launch and then ignore sync forever. Lets also assume that the satellites themselves have such exquisitely precise ISL steering gimbals (well above the requirements for laser communicaitons) and truly outstanding laser pulse timing (for accurate ranging) that each satellite can create an exact model of satellite relative positions independently of any ground stations. (Both of these are extremely generous assumptions that either vastly inflate satellite cost or are plain not physically possible).
Even in that situation, all your GNSS system will tell you is your position relative to that satellite constellation. Not your position relative to Mars, which is what you actually care about - as a gross example, a difference in "what time is it" of 1 minute (a difference of what time the satellite constellation thought Mars had when it was set up vs. the actual local time, not a timing error in the constellation itself) would leave you over 14km away if you were standing at Mars' equator. To solve that, you need to accurately align your constellation to the moving surface of Mars. That needs a minimum of 3 independent ground stations, sufficiently spaced to allow trilateration of sufficient accuracy, and those ground stations need a precisely known relationship to each other (e.g. ground-based surveying or some independent triangulation method). Coordinate system alignment is no simple task at planetary scales where assumptions like "gravity always points straight down" or "the ground does not change length" are not strictly true.
If you then incorporate more realistic assumptions, where clocks drift, satellite positions need to tracked at multiple ground sites to allow accurate triangulation, and satellite orbits continuously shift from surface gravity influence and influence from Phobos and Deimos, the demands on the ground segment only increase.

On Earth, its easy to overlook these baseline requirement because we were producing accurate geographic and geodetic surveys centuries before the first satellite was ever launched. On Mars, we do not have that baseline available. We have satellite imagery of the surface, but that is very much not the same as geodetic survey.


A GNSS system for Mars is obviously not impossible, but it is much more complex than just throwing some clocks on some satellites.

Surely if all the satellites know where they are in relation to each other in X,Y,Z,T (which, as you say, is not trivial) then all you need is a single reference point on the surface that 4 or more sats can reach.

From there, any other point that can be reached by 4 or more sats can be located relative to the reference point, right?

Or am I missing something?

Some poor service would be possible with a single reference point.

A better service needs at least 3 (probably more) reference points.

It depends on what the point of the service is?  If it is a few kilometre accuracy you need, then one reference point is probably fine.  If you want to know the precise location of a rock or for a Seismometers you need high accuracy.   What is the Martian GPS for?

[Brigantine]

Surely if all the satellites know where they are in relation to each other in X,Y,Z,T (which, as you say, is not trivial) then all you need is a single reference point on the surface that 4 or more sats can reach.

From there, any other point that can be reached by 4 or more sats can be located relative to the reference point, right?

Or am I missing something?

Apparently the Chandler wobble is only 10cm - 2 orders of magnitude less than on Earth

[steveleach]

Surely if all the satellites know where they are in relation to each other in X,Y,Z,T (which, as you say, is not trivial) then all you need is a single reference point on the surface that 4 or more sats can reach.

From there, any other point that can be reached by 4 or more sats can be located relative to the reference point, right?

Or am I missing something?

Some poor service would be possible with a single reference point.

A better service needs at least 3 (probably more) reference points.

It depends on what the point of the service is?  If it is a few kilometre accuracy you need, then one reference point is probably fine.  If you want to know the precise location of a rock or for a Seismometers you need high accuracy.   What is the Martian GPS for?

Why do you need more than one ground-based reference point, out of interest?  As far as I can figure, all you need is one point in the ground to fix the planet in relation to the shell of satellites.

[waveney]

Surely if all the satellites know where they are in relation to each other in X,Y,Z,T (which, as you say, is not trivial) then all you need is a single reference point on the surface that 4 or more sats can reach.

From there, any other point that can be reached by 4 or more sats can be located relative to the reference point, right?

Or am I missing something?

Some poor service would be possible with a single reference point.

A better service needs at least 3 (probably more) reference points.

It depends on what the point of the service is?  If it is a few kilometre accuracy you need, then one reference point is probably fine.  If you want to know the precise location of a rock or for a Seismometers you need high accuracy.   What is the Martian GPS for?

Why do you need more than one ground-based reference point, out of interest?  As far as I can figure, all you need is one point in the ground to fix the planet in relation to the shell of satellites.

1) Mars is NOT a perfect sphere. (This will perturb the sats)
2) Mars has moons (these will perturb the sats)
3) The solar system has other bodies (these will perturb the sats)

etc etc.

Finding out the perturbation of the sats is quite useful for mapping gravity anomalies

[steveleach]

Why do you need more than one ground-based reference point, out of interest?  As far as I can figure, all you need is one point in the ground to fix the planet in relation to the shell of satellites.

1) Mars is NOT a perfect sphere. (This will perturb the sats)
2) Mars has moons (these will perturb the sats)
3) The solar system has other bodies (these will perturb the sats)

etc etc.

Finding out the perturbation of the sats is quite useful for mapping gravity anomalies

Yep, but why do you need more than one ground station to handle that perterbation?

Not that it matters, because I've since realised that you need multiple groundstations anyway: you have a sphere within a sphere (sorta) and with just one point on the inner sphere there is still a fair amount of freedom to rotate it slightly.

[DanClemmensen]

Mind you, a GPS satellite at its core is a atom clock (those can be made remarkably small nowadays) that is broadcasting its local time and position. That would not necessarily constitute a trivial modification to Starlink satellites (on top of the bigger solar PV panels), of course.

The clock is only half the solution. The other half is the accurate and up to date measurement of the true orbital parameters of each satellite from a fixed reference (for earth-bound GNSS, that's a reference to fixed ground stations). Orbits will drift and timings will drift - even for atomic clocks - over time, so the system requires constant measurement of those parameters and updating of the satellites that broadcast that up to date orbit reference data along with the timecodes - for Navstar GPS, that would be the Operational Control Segment.

GPS is an incredible achievement, but its basic architecture is 45 years old. Adding explicit positioning capability to Starlink could take advantage of existing Starlink features like the ISL links and much more capable onboard computers. Adding hardware-assisted timestamps (conceptually similar to IEEE 1588 but with sub-nanosecond precision) to the ISL links would increase the satellite position accuracy and clock accuracy for all the satellites, averaging out the uncertainties in the positions as determined by the ground control links.

That doesn't actually solve the problem: Mars has no groundside measurement reference and control system, and Mars has no independent baseline geodetic survey to align to or from.

Lets assume for a moment that our satellites have magical super-atomic clocks that never ever drift and can ignore relativistic effects, such that we do not need to concern ourselves with satellite-to-satellite relative drift, we can synchronise all the clocks at launch and then ignore sync forever. Lets also assume that the satellites themselves have such exquisitely precise ISL steering gimbals (well above the requirements for laser communicaitons) and truly outstanding laser pulse timing (for accurate ranging) that each satellite can create an exact model of satellite relative positions independently of any ground stations. (Both of these are extremely generous assumptions that either vastly inflate satellite cost or are plain not physically possible).
Even in that situation, all your GNSS system will tell you is your position relative to that satellite constellation. Not your position relative to Mars, which is what you actually care about - as a gross example, a difference in "what time is it" of 1 minute (a difference of what time the satellite constellation thought Mars had when it was set up vs. the actual local time, not a timing error in the constellation itself) would leave you over 14km away if you were standing at Mars' equator. To solve that, you need to accurately align your constellation to the moving surface of Mars. That needs a minimum of 3 independent ground stations, sufficiently spaced to allow trilateration of sufficient accuracy, and those ground stations need a precisely known relationship to each other (e.g. ground-based surveying or some independent triangulation method). Coordinate system alignment is no simple task at planetary scales where assumptions like "gravity always points straight down" or "the ground does not change length" are not strictly true.
If you then incorporate more realistic assumptions, where clocks drift, satellite positions need to tracked at multiple ground sites to allow accurate triangulation, and satellite orbits continuously shift from surface gravity influence and influence from Phobos and Deimos, the demands on the ground segment only increase.

On Earth, its easy to overlook these baseline requirement because we were producing accurate geographic and geodetic surveys centuries before the first satellite was ever launched. On Mars, we do not have that baseline available. We have satellite imagery of the surface, but that is very much not the same as geodetic survey.


A GNSS system for Mars is obviously not impossible, but it is much more complex than just throwing some clocks on some satellites.

Of course. You need at least two ground stations, and more are better, but each ground station is not much more than a grounded satellite. You designate one as the Martian base location, like Greenwich on Earth. During initial operation of the system, these stations learn and then refine their locations over a period of weeks as the system calibrates itself. Reference time is a consensus based on all of the clocks, like UTC here on Earth, and would be a lot more stable and accurate than any one of the clocks. It might be helpful to install more accurate clocks at the ground stations, and coordinate them and then slave the satellite's clocks to them, but this is a refinement.

[DanClemmensen]

Lets assume for a moment that our satellites have magical super-atomic clocks that never ever drift and can ignore relativistic effects, such that we do not need to concern ourselves with satellite-to-satellite relative drift, we can synchronise all the clocks at launch and then ignore sync forever. Lets also assume that the satellites themselves have such exquisitely precise ISL steering gimbals (well above the requirements for laser communicaitons) and truly outstanding laser pulse timing (for accurate ranging) that each satellite can create an exact model of satellite relative positions independently of any ground stations. (Both of these are extremely generous assumptions that either vastly inflate satellite cost or are plain not physically possible).

The clocks do not need to be perfect. They will be adjusted to reach a consensus among all the clocks in the network. This requires software, not expensive hardware.

The ISL only needs to be accurate enough to allow intersatellite communication. The  only thing a pair of communicating satellites need on this link is an extremely precise tic embedded in the signal, similar to the IEEE 1588 hardware-supplied tic. One of these every second will be more than adequate. A single tic can be as precise as the baud rate (i.e., the symbol rate, not the bit rate.) I do no know the baud rate for a Starlink ISL but with an aggregate transmission rate of 100Gbps the baud rate is unlikely to be less than 1GHz and the tic is precise to within one nanosecond (30 cm). The two satellites determine their separation by communicating the time of receipt of the tic as measured by their own clock. This requires a very small amount of extra functionality in the transmitter's encoder electronics and the receiver's decoder electronics, plus software.

Note that this can all be done using the "GPS" omnidirectional radio signals instead of the lasers, albeit much more slowly. They already have the equivalent of an extreme precision tic. This may be needed during deployment so the satellites can find each other in the first place.

[CuddlyRocket]

What is the Martian GPS for?

This is the fundamental question, which no-one is answering. People seem lost in the engineering romance of having a Martian GPS. And a Mars-wide communications network. But what's the actual need; who needs it, and when will they need it?

Who's going to design and install these capabilities? And who goes to pay for design, installation and maintenance? Are these the same people who have the need? If not, how to deal with the financial mismatch?

[intelati]

What is the Martian GPS for?

This is the fundamental question, which no-one is answering.... And a Mars-wide communications network. But what's the actual need; who needs it, and when will they need it?

Who's going to design and install these capabilities? And who goes to pay for design, installation and maintenance? Are these the same people who have the need? If not, how to deal with the financial mismatch?

NASA. And honestly they *should be* buying/investigating commercial options.

Their "communication network" consists of "Mars Reconnaissance Orbiter" (a Martian 'Spy Sat') and (a couple? [Currently listening to back episodes of WeMartians, so I forget the State of Mars currently]) other SATs with other uses (and Rover communication takes away from their scientific value)

Hence my recommendation to take that use off of said satellites with Starlinks??!!?

[DanClemmensen]

What is the Martian GPS for?

This is the fundamental question, which no-one is answering. People seem lost in the engineering romance of having a Martian GPS. And a Mars-wide communications network. But what's the actual need; who needs it, and when will they need it?

Who's going to design and install these capabilities? And who goes to pay for design, installation and maintenance? Are these the same people who have the need? If not, how to deal with the financial mismatch?

If humans or robots are going to Mars at all, they need to know where they are and they need to communicate.

GPS is the most cost-effective way to do precision surveying. It is used for this purpose now here on Earth to dramatically increase the efficiency of survey crews.

The comms part of the system ("starlink") can be quite modest: maybe 24 satellites in medium-altitude orbits. This is also exactly the constellation size and orbit that can implement GPS.

There is no need for a bigger constellation until the small one saturates. If it saturates, it will be because the demand is high and there are enough customers to pay for a larger constellation.

[ccdengr]

GPS is the most cost-effective way to do precision surveying.

Certainly, but that doesn't count the $12B the DoD spent on initial deployment and the roughly $2B annual operations cost.

I think people are not appreciating the difficulty of building out a GPS system, though I'm sure it could be done for less than DoD.

[DanClemmensen]

GPS is the most cost-effective way to do precision surveying.

Certainly, but that doesn't count the $12B the DoD spent on initial deployment and the roughly $2B annual operations cost.

I think people are not appreciating the difficulty of building out a GPS system, though I'm sure it could be done for less than DoD.

The first GPS satellite launched 45 years ago. It was bleeding edge and hyper-expensive. Basically everything except the satellite busses and the launches are now cheaper by a factor of 1000 or more. You could almost certainly launch a new-design GPS constellation in a single launch.