How does SpaceX communicate with their vehicles on orbit (and beyond)?

[SkipMorrow]

Do they have permission to use TDRS? Do they have their own tracking and communications stations on earth? What about DSN for things going to Mars-ish orbits?

[SkipMorrow]

An ever so polite and humble bump. Please?

[su27k]

Just go by the stuff I read from here and reddit:

For regular launches SpaceX uses their own ground stations around the world, exact number of stations is unknown but you can hear the stations AOS/LOS being called out during launch webcast.

Dragon uses TDRS.

For Red Dragon, they were planning to rely on NASA for communication, so probably DSN. Not sure what the future plan is regarding BFR.

[speedevil]

For Red Dragon, they were planning to rely on NASA for communication, so probably DSN. Not sure what the future plan is regarding BFR.

Mention has been made of comms satellites around Mars.
Starlink will, in the next couple of weeks launch the first two demonstration satellites.

These satellites have onboard (if they are representative of the fleet) several 6" LASER comms telescopes.

These can in principle, with the right software and hardware choices that may not affect their normal function do at least many kilobits a second from Mars distance. (to 1m telescopes on the ground).

With slight adjustments (larger telescope mainly), it is at least plausible that BFR near (or on) Mars can talk at moderate rates directly to Starlink satellites around earth.

With Starlink derived LASER comms satellites in earth orbit, it can do considerably better than moderate rates.

With several of these larger LASER comms satellites both in earth orbit, and three or four in Mars orbit, you can have 'broadband' up to really quite a large capacity.

This would all be incrementally built on their experience of operating Starlink, and its many many interconnects.

[Jim]

Do they have permission to use TDRS? Do they have their own tracking and communications stations on earth? What about DSN for things going to Mars-ish orbits?

USN

[Jim]

Just go by the stuff I read from here and reddit:

For regular launches SpaceX uses their own ground stations around the world,

Not their own

[jpo234]

Do they have permission to use TDRS? Do they have their own tracking and communications stations on earth? What about DSN for things going to Mars-ish orbits?

USN

A quick Google search revealed this:

http://www.sscspace.com/usn

[flyright]

Do they have permission to use TDRS? Do they have their own tracking and communications stations on earth? What about DSN for things going to Mars-ish orbits?

USN

A quick Google search revealed this:

http://www.sscspace.com/usn

From the link:

Founded as Universal Space Network (USN)

SSC Space US, Inc. provides space operations and ground network services.

[SkipMorrow]

Thanks everyone!

[Steve D]

I imagine when Starlink is operational they could use it for communication with their rockets. Should be plenty of bandwidth and they would never be out of range while in earth orbit.

[Jim]

I imagine when Starlink is operational they could use it for communication with their rockets. Should be plenty of bandwidth and they would never be out of range while in earth orbit.

there are issues using it

[speedevil]

I imagine when Starlink is operational they could use it for communication with their rockets. Should be plenty of bandwidth and they would never be out of range while in earth orbit.

there are issues using it

Which could, if designed in, be non-issues.

Combining the large number of 'spares' that do not need to talk to the ground at any time due to not being over very high traffic areas, the optical links between satellites, ability to share highly dynamically downlink frequencies, and ... - adding a limited 'near earth orbit' capacity may be nearly free.
The normal concerns of radiation patterns of antennas, ... are mitigated if you can simply flip satellites to point directly at your vehicle with no customer impact.

Once they have a nice working satellite bus, that they're making thousands of, adding a couple with a 30cm, not 15cm optical link dish is not hugely expensive.
A little testing, then iterate up to a meter, and you can go quite far indeed.

[AncientU]

Since LEO ops are lower than the Starlink LEO constellation, the launch vehicle is closer to the sats than ground stations/customers, so distance is not the issue.  Beam steering of broadband links should cover any bandwidth needed at any point in the launch.  Interference from the plume should be downward directed, not upward.  There may be interference issues since the launch vehicle is traveling across a wide swath of territory...

Seems the basic pieces for normal communication are there, but many other technical or regulatory issues may be involved.  Would save the cost of ground station fees.

[Robotbeat]

There's no real reason to integrate Starlink into launch ops. Except at Mars.

[AncientU]

There's no real reason to integrate Starlink into launch ops. Except at Mars.

Why keep paying for intermittent and low bandwidth ground stations during orbital operations?  Launch might be different -- especially early ascent needs for redundant telemetry -- but once on orbit, why not if it is technically feasible.

Seems like there is no real reason to not.  Of course, Starlink is a few years into the future...

[Lar]

fits bringing things in house to save expense and increase flexibility.

[Jim]

\
Once they have a nice working satellite bus, that they're making thousands of, adding a couple with a 30cm, not 15cm optical link dish is not hugely expensive.

Yes, it is.  It is thousand times expensive.  Adding it to the spacecraft would be more than the fees paid to ground stations.

[Jim]

The normal concerns of radiation patterns of antennas, ... are mitigated if you can simply flip satellites to point directly at your vehicle with no customer impact.

Which spacecraft and at what time?  Real time planning is not going to be easy. 

[Jim]

Since LEO ops are lower than the Starlink LEO constellation, the launch vehicle is closer to the sats than ground stations/customers, so distance is not the issue.  Beam steering of broadband links should cover any bandwidth needed at any point in the launch.  Interference from the plume should be downward directed, not upward.  There may be interference issues since the launch vehicle is traveling across a wide swath of territory...

Seems the basic pieces for normal communication are there, but many other technical or regulatory issues may be involved.  Would save the cost of ground station fees.

Doppler effects.

[speedevil]

The normal concerns of radiation patterns of antennas, ... are mitigated if you can simply flip satellites to point directly at your vehicle with no customer impact.

Which spacecraft and at what time?  Real time planning is not going to be easy.

The constellation already has real-time planning of spectrum allocated and depends on being able to transfer users between satellites seamlessly, with many satellites being in view.
Dropping one satellite out from servicing earth-bound customers, especially when you can have that satellite be an under-utilised one is not going to affect users.
Even assuming you do in fact need to drop it out.

Similarly, redundancy is needed in the LASER links, and slightly reducing or eliminating the redundancy in one node for optical comms does not seem huge.
If you mean 'how does an external vehicle say it wants comms' - if it's not constantly connected, polling works fine, as does any number of other schemes from dedicated calling frequencies on.

[envy887]

The normal concerns of radiation patterns of antennas, ... are mitigated if you can simply flip satellites to point directly at your vehicle with no customer impact.

Which spacecraft and at what time?  Real time planning is not going to be easy.

Starlink is going to be hitting thousands of moving targets in real time including inter-satellite links.

[speedevil]

Starlink is going to be hitting thousands of moving targets in real time including inter-satellite links.

Millions, unless things go very wrong.

[jpo234]

There might be a market for an in-orbit commercial data relay system for satellites

http://spacenews.com/international-ground-stations-tricky-for-smallsat-operators-to-license/

[john smith 19]

Do they have permission to use TDRS? Do they have their own tracking and communications stations on earth? What about DSN for things going to Mars-ish orbits?

USN

They use the US Navy?

[SkipMorrow]

Do they have permission to use TDRS? Do they have their own tracking and communications stations on earth? What about DSN for things going to Mars-ish orbits?

USN

They use the US Navy?

Good! I wasn't the only one that thought that.

[john smith 19]

Do they have permission to use TDRS? Do they have their own tracking and communications stations on earth? What about DSN for things going to Mars-ish orbits?

USN

They use the US Navy?

Good! I wasn't the only one that thought that.

I'm presuming he meant the DSN, who I thought was run by JPL, but some of the ownership and control pathways around NASA are for historical reasons quite Byzantine.

It's a nice demonstration that when you use low bandwidth communications every character counts and error detection becomes very important.

[ccdengr]

I'm presuming he meant the DSN...

No, he meant Universal Space Network, https://en.wikipedia.org/wiki/Universal_Space_Network

[AncientU]

Since LEO ops are lower than the Starlink LEO constellation, the launch vehicle is closer to the sats than ground stations/customers, so distance is not the issue.  Beam steering of broadband links should cover any bandwidth needed at any point in the launch.  Interference from the plume should be downward directed, not upward.  There may be interference issues since the launch vehicle is traveling across a wide swath of territory...

Seems the basic pieces for normal communication are there, but many other technical or regulatory issues may be involved.  Would save the cost of ground station fees.

Doppler effects.

Intersatellite doppler effects and the challenge of acquisition, pointing, and tracking are required, but relatively trivial adaptations of the tech needed to steer a thousand beams at points on the ground that are moving at 17,000m/h (27,000km/h) relative to the spacecraft, or to conduct inter-spacecraft laser links to long haul data.

Since this tech removes the man from the loop, replaced by great software and state-of-the-art processors, the problem is actually much easier.  Manually scheduling passes and down-links is so last week -- like having a long distance operator route your cell phone calls...

[john smith 19]

I'm presuming he meant the DSN...

No, he meant Universal Space Network, https://en.wikipedia.org/wiki/Universal_Space_Network

Thank you. That is a new fact I did not know before. AFAIK the DSN was run through JPL and AFAIK (possibly from studying the Kistler K1 design) it seemed civilian LV's either used their own private comms or the TDRSS network.