Hacker News Reader

Comment by lxgr | Hacker News Reader

Elon has mentioned a speed of about 7 Mbit/s for an entire cell (which in these systems usually means a circle with a radius of a few hundred kilometers). That's within an order of magnitude of current LEO systems.

The innovation here is that it'll supposedly work with unmodified off-the-shelf phones. (Personally I'm assuming that at least messaging software will have to be modified in order to use the available bandwidth efficiently.)

View on HN

Apple hasn't disclosed the bitrate of their Emergency SOS feature but from FCC filings I believe it uses Globalstar's L-Band at 1610 Mhz for uplink, and I believe the bitrate Globalstar supports in that channel is 9.6 kbps. So almost three orders of magnitude less than the SpaceX system, not one. In addition I believe that the typical achieved bitrate is even less than that in Apple's case due to the lack of a large dedicated antenna. In demos it typically takes many seconds up to several minutes to send a single message, during which time you are directed to continuously point the phone at the satellite for best reception.

The 7 Mbit/s are for an entire cell, i.e. all devices simultaneously transmitting in it, not per device.

Iridium (a comparable system) used to support about 10 Mbit/s per satellite in its first satellite generation, which had 48 spot beams each. The second generation supports about a megabit per modem, but these have larger antennas than a regular satphone.

> Apple doesn't disclose the bitrate of their Emergency SOS feature but it uses Globalstar's L-Band for uplink and I believe the bitrate of that channel is 9.6 kbps.

I'm almost certain Apple doesn't even use the full 9.6 kbit/s modulation scheme for their emergency SOS feature: As you mention, it's way too slow for that (having used it myself – even just transmitting my location takes a couple of seconds, and that's a message of only a few bytes).

Globalstar satellites are bent-pipe relays, so Apple could be using whatever custom modulation and coding scheme they need to stay within the constraints of their power and SNR envelope. With the iPhone's built-in antenna and transmit power (around 2 watts or so, if I remember correctly), I'd be surprised if it was more than a couple hundred bits per second.

Regular (i.e. non-smartphone) satellite messengers are much faster than that and only take a few seconds to transmit and receive messages of hundreds of bytes.

I hope one day Apple can jam some more features in the downstream. Would be cool to broadcast news or weather updates just by pointing your phone at the sky for a few moments. Satellite should be able to blast the signal.

Downsides: no real "cell" infra for Globalstar, and it doesn't work far outside of landmasses.

I'm fairly certain Apple has bigger plans for Globalstar than "only" emergency SOS. They've spent considerable resources securing 85% of the total capacity already!

Regarding coverage, Apple has also already invested considerably, e.g. in ground station expansions in Hawaii and elsewhere. According to Globalstar's coverage map (likely outdated/non-Apple, since Hawaii is not shown as covered yet), most of the Atlantic Ocean is already covered as well:

https://www.globalstar.com/en-us/coverage-maps

I don't see a need for this. It's pretty clear SpaceX's approach can scale to real voice and data service. All you need are more satellites and bigger antennas and you'll have something way better than a one way broadcast like satellite radio or something.

SpaceX has launched the current constellation with Falcon 9; Starship will be able to launch much bigger satellites much faster and much cheaper. And it will work over 100% of the Earth's surface (modulo legal issues) with continuous coverage, as Starlink already does. And SpaceX won't be the only ones doing it.

Only if you're with T-Mobile or with their roaming partner (in US), and same garbage elsewhere.

(Though you could argue that having a provider-agnostic but platform-dependent comms system is trading one problem for another).

Mostly my dream is to drop my cellular provider entirely. I'm living in an increasingly wifi-covered world whenever indoors and I have lots of background-updating offline content, so having some bare minimum connectivity is all I need.

Could go the other way around and go cellular-only for all connectivity, but Canadian telecoms are hell-bent on limiting cellular data to make that impossible for most.

It really ought to be SpaceX's endgame to sign up every carrier, not just one in each country. I bet their partnerships are not permanently exclusive. And again, SpaceX will not be the only ones doing this.

I think that probably works out to a single LTE carrier per spot, which is kinda brilliant, because then you only need to allocate a tiny amount of spectrum for this.

From what I understand the cell is about 13 km².

13km diameter.. which is about 133km².

LTE user count limits per cell is like 500 to 1.5k, and 133km^2 is on par with Liechtenstein(160km^2, 40k population), so like, 7Mbps shared among top 1 - 3.75% of Liechtensteinians? That sounds prestigious...

Anywhere with Liechtenstein's population density would already have cell service, so this wouldn't get used there anyway. This is for the 0.1-0.2% of the US population that doesn't have cell service at all (still prestigious in their own way, but also likely low income).

There are lots of places in the US and Canada (to name a few) where people regularly go for outdoor recreation, or what have you, and there's no cell service anywhere near. This is very applicable to that case, not just where people actually live.

Densely populated areas and even relatively rural areas will have regular cell towers. This is for the few people out of range at any given time, and also outdoors. 1-3.75% sounds plausible I guess? And there are plenty of much less populated areas than Lichtenstein.

Ah, yes, thanks.

Reply on news.ycombinator.com