Hacker News Reader

Starlink V2 will bring very large antennas to Starlink satellites which will enable direct 5G connectivity, but that connectivity will last for a few seconds, with perhaps thirty minutes between connections.

It sounded like temporary connectivity was only initially; once they have a full shell of V2 satellites I would expect constant connectivity.

> Apple has partnered with Globalstar to use geostationary satellites

I haven't read the announcement, but Globalstar's constellation is in LEO.

Yeah but their satellites are basically oversized analogue relays. So bandwidth is bugger all with little capacity to increase it. So Leo or not it's like having CDMA speeds as your cap. Pretty much useless for anything other than a GPS location ping for emergency use.

Indeed. Until a few years ago I was also under the impression that the satellite phone constellations were geostationary or at least geosynchronous. I remember growing up in the 90s hearing about satellite phones and how they had a very noticeable delay, which I assumed meant they were way up there. But perhaps that was just an older generation of constellations, because Globalstar and Iridium (by far the most popular networks in use by consumer-grade devices) are definitely LEO.

inmarsat does sell a handheld satellite phone product (google "inmarsat isatphone") which uses narrow band data channels and handhelds to talk to geostationary satellites. Its coverage is not quite as solid or reliable as iridium, because you can be easily obstructed by a mountain on your south side if you're at latitude 45N or something, and doesn't extend beyond about 70 degrees north, but it's also priced cheaper than iridium for the hardware and the monthly service.

what most people think of as a satellite phone is indeed LEO since iridium has the lion's share of the market.

1400km, compared to 550km for Starlink.

The way technology is headed there is a 110% chance that the phased array tech gets miniaturized and put into a phone. And once the constellation is filled in, you’ll be able to find any tiny slice of sky and talk to a starlink sat that’s flying through it. Indoors is still a problem but who knows, things are always evolving.

You can't take say Moore's Law for chip tech and slap it on to radio tech. They're fundamentally different physical processes. There are limits to radio transmission and reception bandwidth and range due to the basic physics that you can't end run. The inverse square law is a harsh master, for any given level of technology a transmitter 100x closer is going to have a 10,000x advantage however you slice it.

What they might be able to do is expand this to something like limited texting, or maybe down the line even non-realtime voice messaging.

We are still able to have two-way communication with both Voyagers.

Google’s Lyra codec can already get down to 3Kb/s and be reasonably audible. It’s not a stretch to imagine within a few years we’ll be able to push that to below 1Kb/s.

Taking those two things together, I think it’s fairly reasonable to assume at least text and voice are within reach.

Stable diffusion, backwards on the device, targeting a photo. send photos with only the prompt.

Reminiscent of Vernor Vinge's "evocations", as mentioned in A Fire Upon the Deep:

> The screen showed a color image with high resolution. Looking at it carefully, one realized the thing was a poor evocation…. Kjet recognized Owner Limmende and Jan Skrits, her chief of staff, but they looked several years out of style. Ølvira [the ship] was matching old video with the transmitted animation cues. The actual communication channel was less that four thousand bits per second; Central was taking no chances.

> [...]

> The picture was crisp and clear, but when the figures moved it was with cartoonlike awkwardness. And some of the faces belonged to people Kjet knew had been transferred […] The processors here on the Ølvira were taking the narrow-band signal from Fleet Central, fleshing it out with detailed (and out of date) background and evoking the image shown.

Sounds a lot like DALL-E plus a finely tuned training database on both ends could create this sort super-compression if we wanted to. I could imagine having a crisp, real-time (But artificial) 4K video conference at a few kb/s.

Nvidia is already working on this.

The novel I'm quoting was written in 1992. :)

Phased arrays are already in phones that have mm wave antennas ;) usually only 1 dimension though. But it's not inconceivable they will add one capable to reach a satellite. Obviously the number of elements and thus the gain will be lower than the one Ina starlink station. But with slower speeds you don't need as strong a signal.

Sparse array fallacy. You can't make a phased array too small and have high directionality-- you lose all the power in the side lobes. This one's underlying physics.

You can't really shrink a phased array.

If you use smaller antenna elements you must use a higher frequency, and higher frequencies are more attenuated by the earth's atmosphere (generally, when you get to THz there is a window or else visible light wouldn't be able to penetrate) so they have no chance of making it to a satellite.

If you use less elements you lose all your gain in the side lobes, as the other commenter mentioned

> Apple has partnered with Globalstar to use geostationary satellites

globalstar's existing network and licenses have nothing to do with geostationary at all.

please don't comment on things like this if you have fundamental misunderstandings about the nature of different types of mutually incompatible satellite network technology.

If you are going to be pedantic, it would behoove you to use correct capitalization.

it's a well documented fact that typing in all lower case is very slightly faster for persons accustomed to it as a regular practice, which is a completely different thing than fundamentally misunderstanding the technical parameters of a L/S-band LEO vs geostationary data link, if we want to be pedantic about it.

Starlink is not going to be faster than this necessarily. Starlink is also low bandwidth, but has to be compatible with all phones. So they are not able to design hardware in the phone with a better antenna for this purpose as Apple can do.

> You can't just point an antenna at it and have it work. The antenna needs to track the satellite. And it needs to have a constant line of sight free of obstructions.

People already do this when communicating from the ground via amateur satellites on handheld radios. Here the fellow followed the sweep of the sky with his yagi.

https://youtu.be/jmCgIk8eVWw

The speeds Starlink will deliver are 2G speeds/device. They said 2-4Mbps/cell, and a cell is a wide area. I think it will have continuous coverage once the constellation reaches a certain size. AST SpaceMobile is planning 4g or 5g speeds/device with their 40 Gbps capable satellites.

not geosync, LEO.

Globalstar simplex only takes about 200 mW over 2 seconds to deliver a 72 byte message, into an antenna about 0 dBi gain. See the STX3 transmitter module.

That doesn't feel very efficient. 72 bytes is nothing.

72 bytes is more than your reply here.