Starlink maritime coverage map [pdf] | Hacker News Reader

Starlink maritime coverage map [pdf]

api.starlink.com

57 pointsnotpushkin64 comments

This is a nitpick, but the color tones chosen in the map can be very difficult to differentiate. (recall the visual puzzles when two tones separated by a 3rd tone are shown to be actually identical)

I read the map as:

-- There is full coverage around the coasts of N. America, Europe, Australia/NZ, Southern Brazil, Chile right now

-- Q4 2022 will see the whole band of 20 to 60 deg North covered, as well as 33 deg South and below

-- Q1 2023 will see the rest of the entire Earth covered

Do I have that right?

I'm also curious what physical or ground station-related developments happen to enable this? Is it related to enough satellite density to allow transmission satellite-to-satellite to land anywhere, without needing a nearby ground station any more?

Yes, you have it right.

Actually the map stops at about 60 deg S; it's not clear why it's truncated there; it certainly extends much further north of the equator.

> I'm also curious what physical or ground station-related developments happen to enable this?

The v1.0 satellites have no intersat links, so the only way to get coverage with them is to connect to a satellite that is close enough to a functioning ground station. This is the coastal coverage shown.

The v1.5 satellites they are launching right now contain laser links that let them forward incoming packets to other satellites until they reach one that can contact a ground station, but there are not enough of them up there to provide constant coverage anywhere yet. According to this map, by some time in Q4 there will be enough v1.5 sats up to cover the entire mid-latitudes. They get coverage first because the orbits SpaceX are using cause satellites to spend most of their time near these areas. Eventually SpaceX will launch polar sats and just have enough sats up total that they can cover both the poles and the equator.

(The bands of coverage should be symmetric around the equator, so either the map is drawn imprecisely or you got your latitudes wrong.)

Thanks!

I changed the colors, hope this helps: https://imgur.com/a/VvvCx9H

Red/green colorblind people will have trouble reading this version

In that case: https://imgur.com/a/pTUmU11

(according to this palette https://imgur.com/a/lrNT5yZ)

So they're going to go from basically no maritime coverage to having the world's oceans fully covered within the next two quarters? That's pretty impressive, if they can actually pull it off.

Yeah, going from coastal waters only to full coverage seems like a herculean feat to pull off that quickly as I imagine the constraints keeping it to coastal waters at the moment would all need solved.

It’s seems as if current coverage is ground stations, next phase will be supported by constellation laser links, and final coverage stage is more satellites in the necessary latitudes/orbital planes (as they’re currently constrained by F9 payload capacity and desperately need Starship lift capacity to more rapidly deploy the constellation).

Don't most of the satellites in the current constellation not have the laser links needed for intraconstellation communication? So they'd need to replace those too to keep consistent coverage.

Replace, or simply augment with enough additional satellites to provide adequate bandwidth over the oceanic areas.

I can't believe the oceans would have anywhere near the usage you find in-land, so the links between satellites could be relatively small compared to the rest of the system.

I haven't double checked this, but I strongly suspect they're bandwidth limited by the total amount of uplink/downlink they have, not the inter-satellite bandwidth.

Inter satellite links are basically fiber, just without the fiber (instead they just point the laser in the right direction and rely on the fact that space is empty). Up and down has to use valuable radio spectrum, handle interference and atmospheric effects, etc.

Those same satellites will cover land as well, right? I would guess that they are hugely over provisioned for their maritime usecases in support of their main mission.

Yeah to consistently cover land you'll have the same density over the ocean because they're LEO satellites so they're eventually covering the whole surface between the latitudes it flies over due to it's inclination.

You wind up with the same satellite average satellite density across the middle latitudes anyways because the orbits are very low for satellites. Any particular satellite will eventually service every cell (except for those above the inclination of the main constellation iirc those will eventually (tm) be covered by separate satellites in more polar orbits) between 53N and 53S, which is the orbital inclination of the majority of the constellation. Because of this you don't really have any distinction in function between land and sea servicing satellites because any given satellite will be both.

>Don't most of the satellites in the current constellation not have the laser links needed for intraconstellation communication?

Nope, all v1.5 sats have them, so everything launched since January 2021. They experimented with them earlier for polar orbit sats. At this point that's by far the majority of the constellation and either approaching or at critical mass for the mesh. They're probably testing it globally now.

To sibling: they definitely need Starship, but not so much for pure launch numbers as the ability to launch v2.0 at all. F9 fairing isn't big enough. 2.0 will let them shrink cell sizes and improve density and performance which they need too (obviously now more in light of yesterday's announcement). More 1.5 sats wouldn't help as much with that, though the mesh will help relieve ground station chokes as well as ilenabling beyond station service.

Looks like June 2021 actually? [0] From the same page it looks like a bit under half at 44% are v1.5 birds.

[0] https://en.wikipedia.org/wiki/List_of_Starlink_launches

From what I can glean, it looks like about half of the units in orbit are V1.5 with laser links. I know that they had issues getting the links working, but it seems that they have worked the bugs out.

https://en.wikipedia.org/wiki/List_of_Starlink_launches

Maybe the constraints are artificial. From a layman’s perspective it doesn’t seem all that hard. The satellites are already up there, so why would it be any different from providing coverage on land.

It's entirely different.

Right now Starlink is a bent pipe model. The customer's station connects to a satellite which relays data to the closest ground station. Therefore, the satellite must be in contact with both the customer station and a ground station at the same time. This is not possible over oceans.

Newer Starlink satellites also have a satellite to satellite laser link which is intended to allow relay from consumer station -> satellite -> other satellite -> ground station to provide coverage over oceans, but this functionality has not yet been enabled and relies on enough satellite density that each satellite can see a neighbor.

So we get Starlink on our yachts at the same time our Teslas can drive use autonomously to the harbour? Great!

You mean, into the harbor? Yes!

I would take my Tesla to my private jet, we to try it for our yachts so!

I guess we have seen a Tesla crash into a plane already. Don't think it was a jet.

Better not taking any risks, right?

Not so impressive. All they have to do is ping-pong packets from the first satellite, via other marine subscribers' terminals, to other satellites in range of those, until they get to a satellite in range of a shore hub. I.e., "it's only software."

It will mainly depend on having enough marine subscriber terminals in place, scattered around the world on shipping routes, to ping-pong the packets through. Presumably they already have authority to route other subscribers' traffic through "your" terminal.

How is what you just described "not so impressive"? That sounds remarkably impressive.

Routing is impressive. But we do routing already.

Won't work, there are way too many areas where there aren't enough ships for that.

Yet. There are a hell of a lot of ships. They just need for them to be subscribers.

There are a lot of ships, but there are also a lot of places on the ocean where there will never be a second ship close enough to forward your packets (areas of south pacific for instance), and they are still painting those areas as having coverage.

I understand you are enamored with the idea of using ground repeaters. It is, however, not what SpaceX is doing. They have been very open from the start that they intend to move the entire transport layer to orbit, and they have launched more than a thousand satellites capable of doing that with laser links. That's what they are using.

I absolutely do not care one way or the other.

Starlink can get full ocean-wide coverage with software today, or hardware after they have lofted enough of it, their choice. Immediate practicality depends on how fast they deploy marine and aircraft terminals. Full coverage of course depends on having lots of such terminals. They could speed that up if they care to.

Aside from the fact that I would consider that pretty impressive, it's not at all what they're doing. They're using lasers between the satellites to get the data back to/from the ground.

That, of course, requires that they have lasers on all the necessary satellites, a hard hardware problem.

They've got something like 1200 laser equipped satellites in orbit already with 800 of them already operational (the others aren't in final orbits yet I believe). That's about 45% of all the Starlink satellites in orbit. So I would say it's a solved hardware problem.

An alternative approach would be to relay the signal on the surface, perhaps using a fleet of aquatic animals with the same communication system mounted on them.

You know, sharks with frickin' laser beams attached to their heads.

You could imagine floating ground stations to bounce it between sats, but that would be huge amount capital to cover all the oceans.

It's not going to be routed through other marine subscribers' terminals, but instead directly between satellites (using laser links).

It's the difference between a satellite able to relay between two ground locations (e.g. ship and uplink) and series of satellites able to relay directly to each other. In the first case, you need both endpoints to be in same satellite's footprint. In the second, you need to just have any satellite overhead and any satellite in range of an uplink with a few intervening satellites.

That is, that reflects a leap in technology, not an incremental growth from putting more satellites in orbit.

They're going to piss off so many other excruciatingly overpriced and slow providers. I love it

Not like Starlink's Maritime option is particularly cheap at $5K/mo. Probably paltry for cruise ships and megayachts, but sadly out of reach for the average private individual with a boat.

yeah. the big difference is it actually provides useable Internet which makes it not really comparable to anything else satellite based.

Yeah, people acting like this is a market disruptor that's going to kill Garmin or Iridium simply don't know what they're talking about.

We had some discussions regarding the Starlink maritime coverage map a few weeks ago already, you can find it here:

https://news.ycombinator.com/item?id=32018510

Does anyone know what the situation will be for non-commercial use, i.e. sailboats circumnavigating? Will they be able to use the RV dish and pay $150/month, or will they have to go with the $5000/month maritime option?

I think they'll have a more premium product for vessels that want to redistribute connectivity (i.e. cruise ships) and that probably need larger capacity. Or just a bunch of them :)

I'm in that same boat, no pun intended. I've been waiting for the maritime announcement for awhile for my SV. I was very disappointed that they only announced commercial service.

I'd happily pay a huge premium for a gyroscoped dish, that monthly cost tho....

Check out the "starlink on boats" forum on fb. Working now for folks.

Probably the more expensive option because the mid ocean service is going to require using the intra-constellation laser links where the RV is just a roaming dish that can use the normal one hop to ground station method used in the rest of the network. /Maybe/ they'll come out with a cheaper consumer option but who knows, at any rate it will probably be more expensive than the RV version.

On the one hand, you need to use the laser links for this, on the other hand, the satellites over the water aren't going to be busy, so there's not a lot of congestion to want to use pricing to discourage; certainly, there's some congestion near the shore, but there's a lot of shore.

I highly doubt that the laser links will be solely used to provide oceanic coverage for maritime users.

One of the major contraints on adding new users is SpaceX getting approval for and building new ground terminals in areas of high demand. I expect that balancing the load between ground terminals and bypassing congested cables will be a significant use-case for these laser links.

As such, I don't think it is the use of laser links that will determine the pricing of consumer maritime service when it comes.

I think time will be very expensive on them because while over the ocean they'll be dedicated to the more lucrative commercial maritime and the hypothetical HFT low latency routing products. While the satellites are over land they'll be largely free to do whatever they want with the time with their currently announced plans but making hops across the ocean the commercial contracts are going to get top priority because they'll expect it at 5k/month. And as they bring more things on that require the laser links the time is just going to get more precious and by extension expensive.

I think there is very good reason to believe that the limits of the starlink network are ingress/egress and will remain that way due to spectrum availability/licensing (among other reasons).

I absolutely believe that commercial traffic will get priotized, but I highly doubt that maritime traffic will be capable of saturating an intersatelite connection that is designed to provide a backbone for land areas with higher levels of population density.

Would be interested to see this map overlayed with the orbits of their satellites.

So this is the reason for the Australian BIG ASS BATTERY.

Only to power starlink-down-links.

How much BORING company efforts are going on?

Please don't combine random facts into a useless post.

Dont worry, the humanoid robots will also be installing your solar shingles as well.

Now getting a 404 from the link.

I can imagine someone is kicking themselves for accidentally including that PDF in the repo, and removed it when recoginized. they're now sweating bullets hoping that it wasn't pulled before it was removed. oops!

i like to make up stories to pass the time.

Weird, it does work for me.

Here's an archive copy just in case: https://web.archive.org/web/20220711142141/https://api.starl...

I got the same. Three minutes later it suddenly worked. The content was then served by Fastly, so I guess it was their f-up.

Reply on news.ycombinator.com