SpX Mars Communication Constellation?

[speedevil]

Depending on how versatile the phased array hardware is, you might be able to manage it without extra hardware, just special software (since it's probably using some flavor of software-defined radio anyways to begin with). If not, well, surely it's not that hard to add a couple of extra antenna / radio packages.

Many of the things around and on Mars are at very different frequencies to the normal antennas, so a new antenna is probably needed.

[speedevil]

Can we roll GPS into the communication/constellation package?

Does the BFS need to be used for transportation all the way to Mars?    Would it be possible to put a number of satellites on a relatively high orbit, then move to Mars using a SEP drive?  about 5,6 km/s of deltaV, vs 3,9 for geostationnary.  I understand SpaceX hhas some ion drives that would be about right for the need?

And these could leave well ahead of the actual first transit to MArs and be in place when needed....

In principle, a constellation of more than a handful can be used for GPS.
The previous suggestion of being able to receive objects on/around Mars would greatly help in tying the constellations orientation to the ground.
There are several ways to get things to Mars.
Perhaps the most relevant for small early constellations is that if you go into an eliptic mars capture orbit from a hohmann transfer, that can use as little as 800m/s delta-v.
Then aerobraking like done with Mars Global Surveyor can be done over months to get into the right slots - differently on each plane.

I would not be very surprised if they do some form of launch a synod earlier than BFS. (2020, if 2022)

[lamontagne]

In principle, a constellation of more than a handful can be used for GPS.
The previous suggestion of being able to receive objects on/around Mars would greatly help in tying the constellations orientation to the ground.
There are several ways to get things to Mars.
Perhaps the most relevant for small early constellations is that if you go into an eliptic mars capture orbit from a hohmann transfer, that can use as little as 800m/s delta-v.
Then aerobraking like done with Mars Global Surveyor can be done over months to get into the right slots - differently on each plane.

I would not be very surprised if they do some form of launch a synod earlier than BFS. (2020, if 2022)

Ok, interesting.  So if we fuel up a BFS sufficiently to go into an elleptical orbit, it could launch over 100 tonnes of satellites towards Mars?  Discounting 50 tonnes of packaging and propellant to be on the conservative side?
Is that enough for a significant constellation? 200 satellites or so?
And SEP satellites might leave outside Sinods and loiter about waiting for MArs (sorry, forget the name of that type of orbital transfer).

[speedevil]

Ok, interesting.  So if we fuel up a BFR sufficiently to go into an elleptical orbit, it could launch over 100 tonnes of satellites towards Mars?  Discounting 50 tonnes of packaging and propellant o be on the conservative side?
Is that enough for a significant constellation? 200 satellites or so?

In principle a fully refuelled BFS in LEO can throw its nominal payload (150t) at Mars, and return. In a more eccentric orbit, it can throw more.
Starlink satellites are probably a reasonable average mass, some 400kg or so.
200 satellites would be dramatic overkill, and only really needed with a mature colony - tens would be enough for very complete coverage, considering the 'population' density.

[lamontagne]

Ok, interesting.  So if we fuel up a BFR sufficiently to go into an elleptical orbit, it could launch over 100 tonnes of satellites towards Mars?  Discounting 50 tonnes of packaging and propellant o be on the conservative side?
Is that enough for a significant constellation? 200 satellites or so?

In principle a fully refuelled BFS in LEO can throw its nominal payload (150t) at Mars, and return. In a more eccentric orbit, it can throw more.
Starlink satellites are probably a reasonable average mass, some 400kg or so.
200 satellites would be dramatic overkill, and only really needed with a mature colony - tens would be enough for very complete coverage, considering the 'population' density.

Ok, seems like a good idea then.  Anything else that might be interesting to send on that trip?  Perhaps a few 'cheap' landers?  Anything more or less off the shelf that could be sent for useful preliminary exploration?  Probably already too late due to development time.
I guess if we cannot fill the manifest, we might just partly fuel the vehicle.

[Lemurion]

Ground antennas on Mars can't communicate with Earth during Martian night. SpaceX will have satellites in position to relay at night so that everything doesn't go off the grid at Sundown.

[RonM]

Still lose communication during opposition. Need to put a satellite at Earth-Sun L4 or L5. Somewhere to maintain coverage.

[lamontagne]

Still lose communication during opposition. Need to put a satellite at Earth-Sun L4 or L5. Somewhere to maintain coverage.

Could we add that to a Mars delivery of the first constellation?  I expect it would be larger than a Mars constellation vehicle, but it might follow a similar path and use SEP to move itself into place? 

[speedevil]

Still lose communication during opposition. Need to put a satellite at Earth-Sun L4 or L5. Somewhere to maintain coverage.

Could we add that to a Mars delivery of the first constellation?  I expect it would be larger than a Mars constellation vehicle, but it might follow a similar path and use SEP to move itself into place?

A Mars flyby without capture will get you somewhere in the right direction. Scatter two or three, and you can miss the sun.

[biosehnsucht]

Depending on how versatile the phased array hardware is, you might be able to manage it without extra hardware, just special software (since it's probably using some flavor of software-defined radio anyways to begin with). If not, well, surely it's not that hard to add a couple of extra antenna / radio packages.

Many of the things around and on Mars are at very different frequencies to the normal antennas, so a new antenna is probably needed.

Looks like the rovers tend towards UHF (~400Mhz) when talking to the orbiters, though they support X band to talk to Earth. Both should be well within the capabilities of the Starlink phased arrays if using software defined antenna and radio methods.

400Mhz half-wave is about 14 inches, or just a bit over 1/3 meter. If the user terminals are "pizza box sized" then the phased arrays on the Starlink sats are surely big enough. Certainly for quarter-wave, probably for half-wave. The only question really would be sensitivity.

And as I understand it, it's easier to receive lower frequencies with SDR than higher (you get better resolution / sensitivity with less work). If you absolutely need full wave reception, that shouldn't be too hard to add on, but it might need a bigger antenna than the regular Starlink phased array (depending on how big it is).

So I think it could be done with nothing more than software change, but even if it required a hardware change, it would be trivial and easily done in advance if this was a desired feature. It's not like UHF is hard to do.

[speedevil]

Looks like the rovers tend towards UHF (~400Mhz) when talking to the orbiters, though they support X band to talk to Earth. Both should be well within the capabilities of the Starlink phased arrays if using software defined antenna and radio methods.

'SDR' can lead to unwarranted assumptions.

A real-world 15Ghz phased array antenna will  not have a hundred wideband transmitters amplifiers and receivers capable of doing 0-15Ghz, all connected to wideband ADCs/DACs and limited only by what the elements can emit.

This is expensive and has various issues from strong signals in bands you don't care about interfering, to multiplying the amount of CPU you need by several-many.
For example, for the user uplink from 14-14.5GHz, if you implement that with a simple broadband receiver, you have at least 28 times or so data you would with one covering the band of interest. Then it gets worse than this if your antenna subsystem can do some of the adding of individual elements together. In the extreme case, you might end up needing to process over 3000 times the data.

A 'pure' SDR solution will often use a hundred times more power and resources for a task than a more tuned solution, while performing worse.
With a less pure system,  the SDR element only sees some fraction of the signal, with most of what it doesn't care about already rejected. This still allows enormous flexibility.

As you say however - UHF antennas are not very state of the art, and a small additional antenna system and associated electronics on even some of a constellation would be very useful.

[rakaydos]

Bumping for visibility because the topic was raised in another thread.

[Lar]

From a thread where it didn't belong...

OK, should SpaceX install a GPS system with satellites around Mars?  Same with communication satellites.  Some satellites should be installed between earth and Mars at 90 degrees to the sun for communication when Mars is on the other side of the sun.  Maybe they can get NASA to help pay for this if they want to go to Mars also.  Or maybe other countries or companies for rides to Mars.   

I think the answer is yes they should, and I think it likely they will....

[launchwatcher]

From a thread where it didn't belong...

OK, should SpaceX install a GPS system with satellites around Mars?  Same with communication satellites.  Some satellites should be installed between earth and Mars at 90 degrees to the sun for communication when Mars is on the other side of the sun.  Maybe they can get NASA to help pay for this if they want to go to Mars also.  Or maybe other countries or companies for rides to Mars.   

I think the answer is yes they should, and I think it likely they will....

Not sure about the relays " installed between earth and Mars at 90 degrees to the sun ".

I've seen a couple plausible suggestions:

 1) relays at sun-earth or sun-mars L4 or L5
 2) relays in a halo orbit around sun-earth or sun-mars L2 point which takes them out of the earth or mars orbital plane (peeking "up" above the sun) for part of their orbits.

(2) is probably easier to maintain and has a shorter overall signal path.

[geza]

We have learnt, how does a Starlink satellite look like and how does it operate. It is a good time to discuss, how will be this technology applied for Mars?

The simplest idea is to launch a flock of Starlink satellites towards Mars by a Falcon Heavy, very much like they were deployed to LEO by a F9. The satellites separate after TMI and maneuver independently to low Martian orbit afterwards. I am sure that course corrections can be done with the Krypton thrusters. I am less sure about Martian orbit insertion. Obviously, direct insertion into low Martian orbit is impossible with low-thrust engines. I can imagine, however, to maneuver them to a very distant orbit first, and to spiral them down afterwards. I have no idea, whether the available amount of Krypton would be sufficient both for the travel and for the future operation.

Sure, some changes, like larger solar panels, will be necessary. The satellite-to-satellite connection is a must. At least some of the satellites have to be able to communicate with Earth also. This is something substantial, so I can imagine that Earth comm will be handled by very different kind of satellites, or by surface assets.

Does this scheme seem feasible?

[speedevil]

We have learnt, how does a Starlink satellite look like and how does it operate. It is a good time to discuss, how will be this technology applied for Mars?

The simplest idea is to launch a flock of Starlink satellites towards Mars by a Falcon Heavy, very much like they were deployed to LEO by a F9.

From the thread on Starlink tugs
Rough calculations say lightly modified Starlink satellites can get to Mars with around 30 launchable on a falcon 9, in a year from LLO.
With the rather low internet bandwidth on Mars, and if these were at 1500km or so, this may be quite adequate for an initial constellation, nearly unmodified.

They are 1600s ISP, and Elon has stated they can be used as the basis of probes with 250kg of propellant. The solar panels are rather oversized for the 3kW commonly quoted.
This implies they may even work at Mars unmodified other than perhaps some more heaters on critical parts.
Clearly, comms back to earth would need additional systems.

Slow spiral out from LLO to a slow solar transfer orbit, brake gradually into a capture orbit at ~100km*infinity and slowly aerobrake down at a thermal load of less than sunlight at each periareon over several months (Most recent Mars orbiters have done this)

[DAZ]

I do find it amusing that it is assumed that a Starlink like Constellation will be built around Mars and since it is a constellation is automatically hand waived into being a GPS type system.  It is then assumed that this GPS will automatically have the same accuracy and availability of what we have on Earth.  There is very much more to a GPS type system than it just being a constellation.

A GPS satellite essentially only tells you what time it is at the satellite.  If you know precisely enough where the satellite is located this can tell you where you are located in relation to the satellite.  If you know your distance to enough satellites, you then can know what your position is in relation to those satellites.  The Starlink constellation appears to be heavily dependent on the existing GPS satellites networks.  This would allow them to semi-autonomously keep their proper positions in relation to each other.  You could, in theory, use their internal clocks to allow them to roughly determine their position in relation to each other and thus keep their proper orbital positions.  This would allow them to judge their positions in relation to each other with a DOP (Dilution Of Precision) adequate to within hundreds of meters.  To get more accurate than this would require greatly upgraded clocks.  The positions of all the satellites in relation to any position on Mars would continue to drift.  This is exactly what happens to the GPS satellites in Earth orbit.

To correct for this, you need a minimum of one master clock anchored to the planetary surface.  To get a DOP on Mars similar to what you would get on an Earth-based GPS, you will need multiple clocks anchored at multiple points on the surface as you do on earth.  Each of these ground-based anchored clocks, in turn, determines the positions of the GPS satellites.  They then discipline the GPS satellite clocks and gives them their updated positions.  Depending on your minimum acceptable DOP on Mars will dictate how complicated your Mars based GPS system will be.  It is not a simple hand wave of adding a little bit to a Starlink satellite.

Yes, a simple Starlink constellation could probably give you an inherent DOP in the hundreds of meters, but that is not likely to be a very useful level of position accuracy.  It is probably only telling you essentially what you already know, which is your position to within hundreds of meters.  Almost any INS can give you that degree of accuracy.  It’s when you can use a GPS to discipline your INS into the sub-meter range that you get your most benefit from a GPS type system.  A tremendous amount of time, money, effort, and overall resources go into building/maintaining the Earth-based GPS systems.  Look how much time, money, and effort it took the European Union to essentially duplicate with their Galileo system the United States Navstar GPS.

A Mars based GPS is probably going to be very important.  The sooner it would be available the better.  It is not gonna be like ordering Chinese food as in when you order a dozen you get eggroll.  So it’s not going to be a trivial hand wave of when you have a constellation of communication satellites you get GPS.

[speedevil]

I do find it amusing that it is assumed that a Starlink like Constellation will be built around Mars and since it is a constellation is automatically hand waived into being a GPS type system.  It is then assumed that this GPS will automatically have the same accuracy and availability of what we have on Earth.  There is very much more to a GPS type system than it just being a

To a nontrivial degree, GPS is designed as GPS was designed, because it was 1973.
It is now not 1973.
Communication between satellites was not conceived of at that time as being a sane way to solve the various problems of the system, rather than ground control of timing, and computation on the satellites very much wasn't up to much.
Terabit communication between adjacent satellites was the stuff of bad science fiction.
Standards of parts per trillion stability were not off the shelf.

Determination of the relative positions of the constellation to centimeters or better is trivial with a terabit link between satellites.
DOP is almost not a thing, for similar reasons, as this can be good enough (a handful of meters) that it is quite adequate for all solutions other than road following.

Yes, you need to tie it to an absolute position, but almost the only error here is the integrated eccentricity drift of the satellites that is not modeled due to high order gravity and optical torques.
You need quite a modest camera pointed at the next satellite, and an astrometric solver to tie it to a fixed frame.
And one small telescope ties you to the ground - the retroreflector on Insight may be another option.

Or if you are dropping anything on Mars, dropping a stationary node at the same time gives you a good position reference for the constellation.

GPS was designed with ground control and a complex ground side system for good reason in 1973. Not all of those reasons apply.

[freddo411]

We have learnt, how does a Starlink satellite look like and how does it operate. It is a good time to discuss, how will be this technology applied for Mars?

The simplest idea is to launch a flock of Starlink satellites towards Mars by a Falcon Heavy, very much like they were deployed to LEO by a F9. The satellites separate after TMI and maneuver independently to low Martian orbit afterwards. I am sure that course corrections can be done with the Krypton thrusters. I am less sure about Martian orbit insertion. Obviously, direct insertion into low Martian orbit is impossible with low-thrust engines. I can imagine, however, to maneuver them to a very distant orbit first, and to spiral them down afterwards. I have no idea, whether the available amount of Krypton would be sufficient both for the travel and for the future operation.

Sure, some changes, like larger solar panels, will be necessary. The satellite-to-satellite connection is a must. At least some of the satellites have to be able to communicate with Earth also. This is something substantial, so I can imagine that Earth comm will be handled by very different kind of satellites, or by surface assets.

Does this scheme seem feasible?

Yes.   However, I doubt there is any reason to fly a low martian orbit just for comms.    Using a high martian orbit (perhaps Martian centric orbit) you can get essentially global coverage with 3 sats.    You'd give up very low latency, but then again you are probably talking to Earth, ....

[niwax]

We have learnt, how does a Starlink satellite look like and how does it operate. It is a good time to discuss, how will be this technology applied for Mars?

The simplest idea is to launch a flock of Starlink satellites towards Mars by a Falcon Heavy, very much like they were deployed to LEO by a F9. The satellites separate after TMI and maneuver independently to low Martian orbit afterwards. I am sure that course corrections can be done with the Krypton thrusters. I am less sure about Martian orbit insertion. Obviously, direct insertion into low Martian orbit is impossible with low-thrust engines. I can imagine, however, to maneuver them to a very distant orbit first, and to spiral them down afterwards. I have no idea, whether the available amount of Krypton would be sufficient both for the travel and for the future operation.

Sure, some changes, like larger solar panels, will be necessary. The satellite-to-satellite connection is a must. At least some of the satellites have to be able to communicate with Earth also. This is something substantial, so I can imagine that Earth comm will be handled by very different kind of satellites, or by surface assets.

Does this scheme seem feasible?

Yes.   However, I doubt there is any reason to fly a low martian orbit just for comms.    Using a high martian orbit (perhaps Martian centric orbit) you can get essentially global coverage with 3 sats.    You'd give up very low latency, but then again you are probably talking to Earth, ....

The problem with high Martian orbits is that the moons are a lot lower than on Earth, so anything approaching a stationary orbit will be distorted pretty quickly. If there is the option to deploy a medium orbit constellation instead, that might be more useful.