SpaceX Internet Satellite Network: Starlink

[e-FTW]

I have to admit I have not followed all of the developments and discussions: are there any indications about when and if sat-to-sat links will happen?
I ask because it feels like bouncing off ground stations for everything is likely less efficient than sat-to-sat.

 

[mongo]

I have to admit I have not followed all of the developments and discussions: are there any indications about when and if sat-to-sat links will happen?
I ask because it feels like bouncing off ground stations for everything is likely less efficient than sat-to-sat.

Haven't heard of any developments other than the job postings for laser comms a few months ago.
Definitely needed for efficient trans oceanic hops

 

[Cosmacelf]

Sat to sat links aren’t needed for most applications. Marine and cross globe low latency want sat to sat. Both those are minor in comparison to supporting rural access.

 

[bxr140]

And the sooner non-latency critical traffic hits the ground network, the better the space network performs for those who need it.

 

[BioSehnsucht]

One of the original potential money makers was that with functional laser sat to sat links, you could route traffic across the globe faster than the regular fiber optic paths (light in vacuum being faster than light in glass, etc etc). If they solve that, not only does it make supporting terminals in the ocean easier, but it also lets them tap into the lucrative high frequency trading market by providing better latency than the current options for HFT operators. They can also just add to the total global bandwidth, and improve latency across the globe in general for anyone who wants that.

Without that, it still is a easy win against most rural ISP operators as well as providing service to under/unserved areas, but that's only part of the potential market for Starlink.

I can imagine there being multiple tiers of service when they do have inter-sat links. For normal users, in addition to pricing based on bandwidth options, cheaper would drop you into the the nearest ground terminal to use the regular old network, more expensive would go to the nearest ground terminal for your destination (better overall latency and possibly bandwidth). For HFT and similar businesses, an even higher tier that prioritizes their traffic for better QoS at exorbitant (but affordable for HFT) prices. Then any extra bandwidth can be sold in general to ISPs as backbone connectivity at whatever bulk pricing makes sense.

 

[Grendal]

SpaceX is apparently de-orbiting the two prototype Starlink satellites Tintin1 and Tintin 2.

 

[e-FTW]

SpaceX is apparently de-orbiting the two prototype Starlink satellites Tintin1 and Tintin 2.

 

[e-FTW]

Big changes! One more satellite!
Seriously: significantly lower orbits.
https://www.reddit.com/r/Starlink/comments/g3bm3y/spacex_seeks_to_modify_its_kukaband_ngso_license/

 

[Cosmacelf]

Fascinating document. Were they previously thinking of orbiting so low? I wonder what prompted the change.

Also, SpaceX is always ahead of the curve (from the document):

"To help in this effort, SpaceX also provides all of its ephemeris data to other operators via spacetrack.org and it is the first operator to optimize the usefulness of this data by supplementing it with co-variance data, which allows other operators to better predict the trajectories of SpaceX satellites."

 

[Kenz]

Starlink sign me up. Comcast good bye !

Can't come soon enough for me.

 

[e-FTW]

Fascinating document. Were they previously thinking of orbiting so low? I wonder what prompted the change.

I read this allows a much faster passive disposal in the case of failure. But the main advantages are networking related I think: smaller spot areas, lower latency because of reduced distance, and some orbital whizz-bangery.

 

[bxr140]

I wonder what prompted the change

Basically, what @e-FTW said.

--Easier to get a lot of sats on orbit because each sat needs less propellant for orbit raising. (You can can either launch the same number of sats higher or more satellites lower)
--Fully passive deorbit, which is becoming an increasing problem, and you don't have to carry as much propellant on the sat for deorbiting (further improving launch efficiency)
--Less link loss to the ground = higher data rate
--Closer to the ground = less coverage area per sat = lower data demand per sat (fewer users)
--Closer to the ground = smaller coverage area per beam = less users per beam = better service per user (and generally more efficient beam hopping)
--Closer satellites on orbit = easier to make the ISL links work

 

[Cosmacelf]

Basically, what @e-FTW said.

--Easier to get a lot of sats on orbit because each sat needs less propellant for orbit raising. (You can can either launch the same number of sats higher or more satellites lower)
--Fully passive deorbit, which is becoming an increasing problem, and you don't have to carry as much propellant on the sat for deorbiting (further improving launch efficiency)
--Less link loss to the ground = higher data rate
--Closer to the ground = less coverage area per sat = lower data demand per sat (fewer users)
--Closer to the ground = smaller coverage area per beam = less users per beam = better service per user (and generally more efficient beam hopping)
--Closer satellites on orbit = easier to make the ISL links work

OK, maybe I should rephrase my question. Give all these positives, why did SpaceX originally plan to have higher orbits?

 

[mongo]

OK, maybe I should rephrase my question. Give all these positives, why did SpaceX originally plan to have higher orbits?

Higher sats were going to handle the lateral laser links. Lower would beam to upper that would go sideways back to lower to the ground.

 

[bxr140]

OK, maybe I should rephrase my question. Give all these positives, why did SpaceX originally plan to have higher orbits?

Unpopular answer, but: Unchecked ambition. The Starlink sats are small, which limits both the aperture sizes and the available power. That's plausibly overcome with 40k satellites, but with many fewer satellites--clearly the reality of the next many years--they will have an impossible time meeting any kind of service standard they've envisioned.

To be fair, this is old news. I don't recall when we first heard that they'd drop altitude (is it week 3 of stay-at-home? week 30?) but I'm pretty sure it was months ago. But no mistake, filing 'for a constellation' in the first place is huge and this kind of change is equally huge and not something you do if you're just shuffling around the way you deploy the constellation.

 

[Grendal]

What's going on with Starlink satellites are nestled in this article.
Falcon 9 to become U.S. rocket leader; Starlink “where are they now” edition - NASASpaceFlight.com

 

[e-FTW]

Unpopular answer, but: Unchecked ambition. The Starlink sats are small, which limits both the aperture sizes and the available power. That's plausibly overcome with 40k satellites, but with many fewer satellites--clearly the reality of the next many years--they will have an impossible time meeting any kind of service standard they've envisioned.

The first orbital tests of Starship may accelerate all of this in the next year or so.

 

[bxr140]

The first orbital tests of Starship may accelerate all of this in the next year or so.

Maybe.

Even in the hypothetical era of a fully operational 100T+ two stage Starship (which I'd wager is at least 3 years away), its hard to imagine 40k satellites being put up in any less than 4-5 more years. Layer on the likely need for replenishment sats by that point and IMHO you're looking at the better part of 10 years before a true 40k satellite constellation could possibly be realized.

Somewhat tangent, its hard for me to imagine anywhere near 40k satellites in the constellation ever. I'd wager it doesn't get above a few thousand and definitely not over 10k, and I'd accept that maybe the constellation is fully armed and operational in 5 years from now.

 

[e-FTW]

Maybe.

Even in the hypothetical era of a fully operational 100T+ two stage Starship (which I'd wager is at least 3 years away), its hard to imagine 40k satellites being put up in any less than 4-5 more years. Layer on the likely need for replenishment sats by that point and IMHO you're looking at the better part of 10 years before a true 40k satellite constellation could possibly be realized.

Somewhat tangent, its hard for me to imagine anywhere near 40k satellites in the constellation ever. I'd wager it doesn't get above a few thousand and definitely not over 10k, and I'd accept that maybe the constellation is fully armed and operational in 5 years from now.

Am sure Elon would be like “You’re on! And we will have 40k in half that time!”
What will actually happen, who knows. Let’s come back here in 2025!

 

[e-FTW]

I liked this animation: https://www.reddit.com/r/Starlink/comments/g5odmg/starlink_deployment_animation