Starlink : Uses other than LEO Communications

[speedevil]

edit/gongora: Split off from the Starlink General Discussion - Thread 2.  Please use this thread for discussing future uses for Starlink sats other than the LEO communications constellation that is being deployed now.

I think they’ll do continuous improvement, but lock down each flight load of 72 satellites. Improved chipsets or solar panels or improved mechanicals and engines. By the time the constellation is complete the satellites will be completely upgraded, while possibly never undergoing a single major refresh.

The entire constellation is going to be refreshed approximately every 5 years, as that is the nominal orbital life of an individual satellite.  Whether the satellites themselves get upgraded as frequently is another question; I expect they will.

I idly wonder if the line will be specced for >starlink rates, for the non-comms part of the payload. (panels, GNC, engine).

Starlink tugs and probes to every single >500m near earth asteroid - as potentially interesting options, as well as Starlink-Mars and Moon.

When SS starts flying regularly, the cost of flying these probes would fall to near zero.

I note also the threads:
Starlink derived debris deorbiting and Starlink considered as tugs

[ChrisWilson68]

Starlink tugs and probes to every single >500m near earth asteroid - as potentially interesting options

That's one way to answer the question "Will the U.S. have control over Musk's Mars colony?".

[freddo411]

I think they’ll do continuous improvement, but lock down each flight load of 72 satellites. Improved chipsets or solar panels or improved mechanicals and engines. By the time the constellation is complete the satellites will be completely upgraded, while possibly never undergoing a single major refresh.

The entire constellation is going to be refreshed approximately every 5 years, as that is the nominal orbital life of an individual satellite.  Whether the satellites themselves get upgraded as frequently is another question; I expect they will.

I idly wonder if the line will be specced for >starlink rates, for the non-comms part of the payload. (panels, GNC, engine).

Starlink tugs and probes to every single >500m near earth asteroid - as potentially interesting options, as well as Starlink-Mars and Moon.

When SS starts flying regularly, the cost of flying these probes would fall to near zero.

People have floated the idea that each sat is under a million bucks to make (based upon their own guesses).

I'm drooling with excitement at the thought of a FH throwing 20 or so of these starlink birds with a cameras attached to escape velocity for 20 (simultaneous!) asteroid prospecting missions.    I could imagine a mesh radio network between the  probes for returning data to Earth without using NASA's DSN.

How's that for a spectacular discovery class planetary mission?

[ChrisWilson68]

I think they’ll do continuous improvement, but lock down each flight load of 72 satellites. Improved chipsets or solar panels or improved mechanicals and engines. By the time the constellation is complete the satellites will be completely upgraded, while possibly never undergoing a single major refresh.

The entire constellation is going to be refreshed approximately every 5 years, as that is the nominal orbital life of an individual satellite.  Whether the satellites themselves get upgraded as frequently is another question; I expect they will.

I idly wonder if the line will be specced for >starlink rates, for the non-comms part of the payload. (panels, GNC, engine).

Starlink tugs and probes to every single >500m near earth asteroid - as potentially interesting options, as well as Starlink-Mars and Moon.

When SS starts flying regularly, the cost of flying these probes would fall to near zero.

People have floated the idea that each sat is under a million bucks to make (based upon their own guesses).

I'm drooling with excitement at the thought of a FH throwing 20 or so of these starlink birds with a cameras attached to escape velocity for 20 (simultaneous!) asteroid prospecting missions.    I could imagine a mesh radio network between the  probes for returning data to Earth without using NASA's DSN.

How's that for a spectacular discovery class planetary mission?

It wouldn't work.  Starlink satellites aren't designed to be able to communicate at those kinds of distances.  They're designed to communicate at ranges of a few hundred kilometers.  Your mesh network would need millions of satellites.

They're also not designed to have and point a camera.  And their solar panels aren't designed for distances farther than Earth orbit.

There's a reason that planetary probes are purpose-designed rather than taking some existing communication satellite and slapping a camera on it.

[Barley]

People have floated the idea that each sat is under a million bucks to make (based upon their own guesses).

I'm drooling with excitement at the thought of a FH throwing 20 or so of these starlink birds with a cameras attached to escape velocity for 20 (simultaneous!) asteroid prospecting missions.    I could imagine a mesh radio network between the  probes for returning data to Earth without using NASA's DSN.

How's that for a spectacular discovery class planetary mission?

You are underestimating the size of the mesh.  Hayabusa-2 uses a 90cm dish to communicate with a 34m DSN dish at a distance of about 2AU.   Ceteris paribus* the range between two 90cm dishes would be reduced by 0.9/34 ~1/38 (or about 0.05AU).  So you'd need at least 38 perfectly placed relays.   Orbital mechanics means you can't have perfectly placed relays.  At a guess you'd need something like 60 relays to maintain continual communications over several years.  Almost all of them in deep space near nothing of interest.

OTOH throw up 100 relays and you could probably support missions to most near Earth asteroids.  Or 500 relays and hundreds of probes to large parts of the main belt.  Interesting, but no longer a discovery class mission.

A mesh network would be a rather large piece of infrastructure.  Unless it is supporting many missions to the same region it would be better to use the DSN.

* Ceteris non paribus.  The DSN is much higher power and has better sensitivity than the spacecraft.  I'd guess a another factor of at least 2 is needed here.

[speedevil]

It wouldn't work.  Starlink satellites aren't designed to be able to communicate at those kinds of distances.  They're designed to communicate at ranges of a few hundred kilometers.  Your mesh network would need millions of satellites.

They're also not designed to have and point a camera.  And their solar panels aren't designed for distances farther than Earth orbit.

There's a reason that planetary probes are purpose-designed rather than taking some existing communication satellite and slapping a camera on it.

Certainly.
However, orbiting at 1km, enabled by a capable ion engine rather than a flyby mission rather eases nearly all constraints on the camera.
And yes, of course power and propulsion are not the only parts of the spacecraft that matter.

[freddo411]

People have floated the idea that each sat is under a million bucks to make (based upon their own guesses).

I'm drooling with excitement at the thought of a FH throwing 20 or so of these starlink birds with a cameras attached to escape velocity for 20 (simultaneous!) asteroid prospecting missions.    I could imagine a mesh radio network between the  probes for returning data to Earth without using NASA's DSN.

How's that for a spectacular discovery class planetary mission?

You are underestimating the size of the mesh.  Hayabusa-2 uses a 90cm dish to communicate with a 34m DSN dish at a distance of about 2AU.   Ceteris paribus* the range between two 90cm dishes would be reduced by 0.9/34 ~1/38 (or about 0.05AU).  So you'd need at least 38 perfectly placed relays.   Orbital mechanics means you can't have perfectly placed relays.  At a guess you'd need something like 60 relays to maintain continual communications over several years.  Almost all of them in deep space near nothing of interest.

OTOH throw up 100 relays and you could probably support missions to most near Earth asteroids.  Or 500 relays and hundreds of probes to large parts of the main belt.  Interesting, but no longer a discovery class mission.

A mesh network would be a rather large piece of infrastructure.  Unless it is supporting many missions to the same region it would be better to use the DSN.

* Ceteris non paribus.  The DSN is much higher power and has better sensitivity than the spacecraft.  I'd guess a another factor of at least 2 is needed here.

Thanks for doing some math on this.    I wasn't able to find any detailed specs on the starlink birds.    I'll speculate and say that their antenna might be a factor of 4 bigger, and their available power is probably roughly the same as Hayabusa-2 (2.6 kW).    Looks like the mesh idea is plausible, but would be more realistic with better than standard starlink hardware

[cro-magnon gramps]

Hey Freddo, My own viewpoint, is that the Starlink Operations is a cover for more than just the present Starlink v.1 Earth Satellites... It is accepted that those satellites will iterate in complexity over the lifetime of the Earth Operations.
The idea of further uses for the Starlink Operations opens the door to various styles and uses for satellites, with increasing complexity, but built along the same ideas on an assembly line basis.

Whether it is Deep Space Exploration, as you have mentioned, or a Space Network for communications between planets, or a Mars Constellation of Satellites, all are possible under the umbrella of Starlink Operations. Like a Super Nova, the expanding field is limited only by the strength of the initial explosion...

In 50 years who knows what Starlink will or can become? I've been excited by it and Starship ever since they were named. 

[Ludus]

Hey Freddo, My own viewpoint, is that the Starlink Operations is a cover for more than just the present Starlink v.1 Earth Satellites... It is accepted that those satellites will iterate in complexity over the lifetime of the Earth Operations.
The idea of further uses for the Starlink Operations opens the door to various styles and uses for satellites, with increasing complexity, but built along the same ideas on an assembly line basis.

Whether it is Deep Space Exploration, as you have mentioned, or a Space Network for communications between planets, or a Mars Constellation of Satellites, all are possible under the umbrella of Starlink Operations. Like a Super Nova, the expanding field is limited only by the strength of the initial explosion...

In 50 years who knows what Starlink will or can become? I've been excited by it and Starship ever since they were named.

They’ve talked about selling the standard mass produced Starlink components like folding PV panels, batteries, propulsion, as a platform for other spacecraft. The non-LEGOness of spacecraft is not a law of nature, it’s a problem to be overcome. Somebody needs to produce a deep space communications plugin, then the platform can be sent in quantity to visit/prospect asteroids.

[Cheapchips]

They’ve talked about selling the standard mass produced Starlink components like folding PV panels, batteries, propulsion, as a platform for other spacecraft. The non-LEGOness of spacecraft is not a law of nature, it’s a problem to be overcome. Somebody needs to produce a deep space communications plugin, then the platform can be sent in quantity to visit/prospect asteroids.

Have they talked about selling components?  I thought they're only talked about other uses?

I don't think that SpaceX will go the components route unless they've explicitly said it.  They'll sell 'slots' on a Starlink payload version.  Design your experiment/service to interface with the Sat's power and comms.  That's it.  They'll go up on a rideshare flight. You might even be a satshare, with other customers in another slot. As a customer, you're never a satellite operator.

Hopefully Musk will talk more about further uses for Starlink at March's SatShow keynote.  SpaceX are making more sats per day than they need for biweekly constellation launches. They should have 'loads' of spares knocking about.

[Cherokee43v6]

Question out of curiosity.

Using the principles of interferometry(sp?) (ie using two or more widely separated telescopes to resolve a clearer image) could a system like Starlink be used as a deep space communications receiver and could this be a current (or future) secondary mission for Starlink in line with Elon's Mars goals?

[Eerie]

Question out of curiosity.

Using the principles of interferometry(sp?) (ie using two or more widely separated telescopes to resolve a clearer image) could a system like Starlink be used as a deep space communications receiver and could this be a current (or future) secondary mission for Starlink in line with Elon's Mars goals?

No.

[Redclaws]

Eerie,

Can you talk a little more about why?  I think it’s an interesting question and would be curious to hear the reasons.

[Semmel]

Eerie,

Can you talk a little more about why?  I think it’s an interesting question and would be curious to hear the reasons.

Radio interferometry works by storing the analogue signal on all telescopes with excruciating precision and time stamps. Later, the signals of all telescopes are combined in a supercomputer to produce interference digitally by combining different signals.

Starlink sats are used to process digital data. The first thing these antennas do is throw the analogue signal away and extract the digital signals. All routers and/or repeaters, telecom antennas, etc work like that. Doing interference astronomy is very, very different.

[OTV Booster]

Question out of curiosity.

Using the principles of interferometry(sp?) (ie using two or more widely separated telescopes to resolve a clearer image) could a system like Starlink be used as a deep space communications receiver and could this be a current (or future) secondary mission for Starlink in line with Elon's Mars goals?

Interferometry can increase resolution but the signal strength reaching both (all) collectors adds together for total signal strength. This holds true at all wavelengths. To pick up a weak signal total collector area is what counts.

[Tommyboy]

Eerie,

Can you talk a little more about why?  I think it’s an interesting question and would be curious to hear the reasons.

As I understand it, you need to know the distance between your two receiver to within a fraction of the wavelength you are capturing (so we're talking about (tens of) nanometers), and as other people said you need to add the analog signals together.
We aren't yet capable of doing that in space, but there are ongoing experiments.

Edit: Disclaimer: I'm just an Electrical Engineering drop-out now working in cybersecurity. I could be wrong.

[archae86]

To pick up a weak signal total collector area is what counts.

Look up the Deep Space Network (DSN) facilities to see what you get when you design for the purpose of talking to gadgets at planetary distances.

70-meter steerable antennas point right at the object of interest.  The size and pointing right at the object means lots of signal collected.  The geometric gain means random noise is way down compared to an equal area flat antenna.  Very, very low-noise receivers hold down the component of noise coming inside the receiving equipment itself.

Signal to noise ratio is nearly the entire game.  If you integrate the frequency-specific S/N over the frequency range in use and do just a little math, you get the Shannon limit for the maximum data rate possible to transmit.

For openers, Starlink antennas are pointed the wrong way, they are not very big (compared to 70 meters) and they don't have the geometric gain required to lower background noise.  I doubt the receiver electronics are nearly so low-noise as DSN either.

However... DSN has just three antennas distributed around the earth, so it is a very oversubscribed resource.  Someone is going to suggest considering putting much higher power transmitters on some planetary-distance clients, so that much smaller (cheaper) facilities can handle much of the communication load.

But I don't think those smaller and cheaper facilities will be a secondary function of the Starlink fleet.

[meberbs]

Eerie,

Can you talk a little more about why?  I think it’s an interesting question and would be curious to hear the reasons.

As I understand it, you need to know the distance between your two receiver to within a fraction of the wavelength you are capturing (so we're talking about (tens of) nanometers), and as other people said you need to add the analog signals together.
We aren't yet capable of doing that in space, but there are ongoing experiments.

Edit: Disclaimer: I'm just an Electrical Engineering drop-out now working in cybersecurity. I could be wrong.

You were almost right, until you got to the part about tens of nanometers. For something like Ka band, the distance is on the order of 1 cm and this could be done if needed. Actually there are ways to get to the order of nanometers in space but that is more difficult, typically low TRL, and would only be used for certain applications, such as optical wavelength interferometry for observations of exoplanets.

[Tommyboy]

Eerie,

Can you talk a little more about why?  I think it’s an interesting question and would be curious to hear the reasons.

As I understand it, you need to know the distance between your two receiver to within a fraction of the wavelength you are capturing (so we're talking about (tens of) nanometers), and as other people said you need to add the analog signals together.
We aren't yet capable of doing that in space, but there are ongoing experiments.

Edit: Disclaimer: I'm just an Electrical Engineering drop-out now working in cybersecurity. I could be wrong.

You were almost right, until you got to the part about tens of nanometers. For something like Ka band, the distance is on the order of 1 cm and this could be done if needed. Actually there are ways to get to the order of nanometers in space but that is more difficult, typically low TRL, and would only be used for certain applications, such as optical wavelength interferometry for observations of exoplanets.

You're right, I read the initial question wrong. I was thinking we were still talking about interferometry in optical wavelengths, instead of "regular" radio. I shouldn't be trying to be a smart-ass in the middle of the night.

[Cherokee43v6]

Thank you everyone who replied.

So it sounds like the concept is 'plausible but unlikely at the current tech level' and thus unlikely to be something incorporated into the current Starlink configuration.

My reason for asking the question is based off of the obvious synergies between all of Elon's current companies in relation to his announced Mars Colonization goals.

SpaceX, Starship obviously.
Tesla, ruggedized electric vehicles would be the most useful form of surface transport; plus the former 'Solar City' products (power wall and solar panels) for power generation and storage.

In light of that, it seemed to me that the addition of the communications infrastructure presented by Starlink could be a key feature if it eventually developed to include the mentioned capability.  Particularly if such a system were put into place around Mars as well.

My guess is that this type of thing will be something to keep an eye out for as Elon's plans progress, as said communications infrastructure would be a minimum requirement to help maintain a 'Martian Technocracy' as Elon phrased it.