Sounds (and is) cool, but it breaks the whole concept of cellular, severely limiting system capacity. Hard to see it scaling up to provide voice for many simultaneous users.
I wonder if they did any SAR level testing on those phones. I mean I’m sure that cell phone has to use a lot of power to transmit to a satellite and I know from FCC regulations you can’t be closer than 11 inches from a stink dish when it’s in operation. I know the two are not the same just to comparison. Of power.
According to https://direct.starlink.com/ they are unmodified phones. I guess that the battery drain will be huge. I don't know how low those satellites will fly but the satellite to phone distance should be in the hundreds of km, not single digit km at most like we usually do.
Communication is bidirectional so there is no need for the downlink to be more powerful than the uplink: the phone must be able to send data to the satellite using its own antenna and battery.
However in theory a more powerful downlink could send data down to the phone faster than the phone sends it up to the satellite.
A really powerful satellite has a power supply of about 1 kilowatt.
Assuming perfectly efficient antennas and a single spot beam with a diameter of about 300 kilometers (realistically, it's going to project dozens of these for a total of several thousand kilometers, but let's stick with the worst case here), that's an area of about 200 000 square kilometers to cover with that one kilowatt, i.e. 5 milliwatt per square kilometer.
This is roughly the same amount of maximum EM radiation emitted by a single AirPod – now spread that over one square kilometer.
Compare that with a regular LTE or 5G cell with a cell radius of at absolute best 30 kilometers (but often much, much less, e.g. 1 kilometer) and an unusually high EIRP of 50 watt, and you get 17 milliwatt per square kilometer. Or take an AM radio station: These transmit with 50 kilowatt and are much closer to people, sometimes located on rooftops in densely populated cities.
Realistically, what arrives down on earth is probably still a few orders of magnitude less than that and just at the edge of the noise threshhold for the phone. It's not exactly a beam weapon.
The phones are normal phones but it's very likely they will be transmitting at their maximum power. Which is the power they're already being tested for.
However because it's only for texting for now, you won't be holding it too close to your head. And it will transmit for only a fraction of a second. Because radio energy drops really fast with distance it won't be a problem and will in fact be better than when you're having a phone call in bad coverage.
Isn't an Inmarsat phone with antenna, like, 8x heavier and 4x more expensive than a not-very-fancy 5G smartphone? A SMS limited to 160 characters for half a dollar?
Standard 5G is a lot more efficient. The difference seems to me at least the same as between a steam train and an electric train.
Apparently it does not support the non-terrestrial networks (NTN) 5G standards [1],[2]. Perhaps it is not a bad thing since 5G phones with NTN capability are just being released but this will make it not future proof. Since it is not Starlink main business of providing satellite cellular services, it should be fine.
[1] Starlink indicates its satellite-to-phone service will drop next year:
With the way the constellation works (needing regular replenishment of satellites as they deorbit every 5 years or so), not currently supporting NTN isn't a big deal. If it sufficiently catches on, the satellites being launched will be given the necessary upgrades and over time the entire constellation will end up with the necessary upgrades.
Most folks commenting seem to miss the fact this is for text messages only, and likely "emergency" ones, i.e. rate limited, at that. This would not result in power drain during normal LTE operation, but - my guess - would instead check with the satellite to send/recv txt messages every now and then AND only when not in LTE coverage.
These satellites mimic 5G so most existing 5G phones should be compatible without additional hardware.
The beams are REALLY low bandwidth (7MB per beam and the beams are “very big”), and that bandwidth is shared with the whole beam. You will only use this for emergencies, at least for now.
The point will be to expand coverage in areas that it may be difficult to add coverage. It won’t replace or augment regular towers.
It will only allow texting later this year, and they expect to scale up to voice and data next year as well as IOT - whatever that means since they broke it out separately.
There is a lot of IoT: autonomous forklifts in a parking lot, TVs that upload which TV shows you watched, oil pumps in the middle of nowhere, smart meters. Some make more sense, others less. But yeah, often consumer privacy and user control of the devices they bought is not the priority of the IoT activated device manufacturers.
Just being able to text and send location would be huge for emergency communications and letting people know you are ok. This will probably completely replace sat based communicators like Garmin inReach.
Weird to see a link to Slashdot these days. My immediate reaction is that it's either a historical article or news they're shutting down. I used to read it daily, much like HN.
5G is a godsend for them because it requires many small cells.
On the other hand, that also enhances the accuracy of things like GPS. I feel like if you combine Galileo + 5G + WiFi sourcing you might get ridiculously accurate.
Only if they knew the number I guess and could ping it and look what transmitter answered. After multiple pings you could probably narrow it down a lot from all the other transmitters. But then again, if they know your number there might just be some Google or Apple API responding with a GPS reading. E.g. Google stores most peoples position anyway in their Maps SaaS.
But with this every shady US authority can map each phone number to a location. I assume that the cell phone reports its number to a tower or at least some unique id?
I really hope "Linux phones" starts to become practical so we can opt-out of all these shenanigans. And I am not even on a Washington shit list.
As I see it, ASTS is perhaps ahead on the antenna technology demonstration, and SpaceX is definitely ahead on deployment at scale (i.e. SpaceX launched six satellites just now, ASTS has launched one afaik). Which of those metrics is "more important" isn't clear to me, but I suspect SpaceX has the overall velocity advantage.
One thing I don't know is how the size of the relevant satellites compares. If ASTS is getting twice the bandwidth from a comparable size satellite, that's impressive. If their satellite is an order of magnitude larger than a Starlink sat that'll be an issue.
I was speaking technologically. Both Starlink and ASTS satellites are technological demonstrations so I would argue only the technology matters. Since ASTS plans to launch using SpaceX and manufacture through 3rd-party contracts I don't believe Starlink's existing constellation provides any benefits; rather the competition will come down to funding and manufacturing expertise.
The 6-1 satellite advantage doesn't necessarily favor Starlink. BlueWalker 3 was launched 15 months before Starlink's first mobile communications satellite and was itself preceded 5 months by BlueWalker 1. However the six Starlink v2-Mini-D2D satellites feature inter-satellite links, so that is a point in favor of Starlink. BlueWalker will be succeeded by five commercial satellites called BlueBirds which are currently under construction.
As for technology, I can't find any information about the Starlink direct-to-cell satellites. As far as I can tell they haven't been even given NORAD IDs yet.The entire Jan 2 Group 7-9 launch seems to be classified as 58699/Object-A. Given that the launch lifted 21 rather than the standard 22 satellites we can assume the new design is similar to the Starlink v2-Mini in size and weight - perhaps only 125 kg heavier. This suggest the efficiency/weight ratio favors Starlink. Starlink v2-Mini weighs 800 kg
1. I don't have access to the FCC filings so I multiplied the new bus length by the old width.
2. Per-beam, I assume.
From what I gather satellite mass is more correlated with coverage (number of beams) rather than beam efficiency. For the moment, ASTS antennas are just better ($/bandwidth not published). And the new BlueBird satellites are going to be horribly massive and horribly bright compared to BlueWalker-3.
P.S. Anyone know what BlueWalker-3-deb represents?
47 comments
[ 5.3 ms ] story [ 76.7 ms ] threadHowever in theory a more powerful downlink could send data down to the phone faster than the phone sends it up to the satellite.
Assuming perfectly efficient antennas and a single spot beam with a diameter of about 300 kilometers (realistically, it's going to project dozens of these for a total of several thousand kilometers, but let's stick with the worst case here), that's an area of about 200 000 square kilometers to cover with that one kilowatt, i.e. 5 milliwatt per square kilometer.
This is roughly the same amount of maximum EM radiation emitted by a single AirPod – now spread that over one square kilometer.
Compare that with a regular LTE or 5G cell with a cell radius of at absolute best 30 kilometers (but often much, much less, e.g. 1 kilometer) and an unusually high EIRP of 50 watt, and you get 17 milliwatt per square kilometer. Or take an AM radio station: These transmit with 50 kilowatt and are much closer to people, sometimes located on rooftops in densely populated cities.
Realistically, what arrives down on earth is probably still a few orders of magnitude less than that and just at the edge of the noise threshhold for the phone. It's not exactly a beam weapon.
However because it's only for texting for now, you won't be holding it too close to your head. And it will transmit for only a fraction of a second. Because radio energy drops really fast with distance it won't be a problem and will in fact be better than when you're having a phone call in bad coverage.
Standard 5G is a lot more efficient. The difference seems to me at least the same as between a steam train and an electric train.
[1] Starlink indicates its satellite-to-phone service will drop next year:
https://www.telecoms.com/telecoms-infrastructure/starlink-in...
[2] Connecting the world with 5G NTN:
https://www.rohde-schwarz.com/se/solutions/test-and-measurem...
* https://spacenews.com/spacex-deploys-direct-to-smartphone-sa...
Perhaps that would be a better link.
These satellites mimic 5G so most existing 5G phones should be compatible without additional hardware.
The beams are REALLY low bandwidth (7MB per beam and the beams are “very big”), and that bandwidth is shared with the whole beam. You will only use this for emergencies, at least for now.
The point will be to expand coverage in areas that it may be difficult to add coverage. It won’t replace or augment regular towers.
It will only allow texting later this year, and they expect to scale up to voice and data next year as well as IOT - whatever that means since they broke it out separately.
But yeah, maybe it'll start consuming so much energy our alien gods will just shut our universe down. Thanks, Elon.
On the other hand, that also enhances the accuracy of things like GPS. I feel like if you combine Galileo + 5G + WiFi sourcing you might get ridiculously accurate.
Phones have been using tower location for more accurate and rapid location for decades. Some of the first standards came in 3G networks.
And the providers in my country have complained about having to place many more towers for 5G, so yeah.
It’s very easy to follow the logic - denser cells allow for more accurate triangulation.
But with this every shady US authority can map each phone number to a location. I assume that the cell phone reports its number to a tower or at least some unique id?
I really hope "Linux phones" starts to become practical so we can opt-out of all these shenanigans. And I am not even on a Washington shit list.
AST SpaceMobile is getting ready for commercial ops this year with „full broadband“ 5G satellites.
For the record, I am not affiliated with them. I just need that kind of service desperately.
https://twitter.com/AST_SpaceMobile/status/17434215614224303...
One thing I don't know is how the size of the relevant satellites compares. If ASTS is getting twice the bandwidth from a comparable size satellite, that's impressive. If their satellite is an order of magnitude larger than a Starlink sat that'll be an issue.
The 6-1 satellite advantage doesn't necessarily favor Starlink. BlueWalker 3 was launched 15 months before Starlink's first mobile communications satellite and was itself preceded 5 months by BlueWalker 1. However the six Starlink v2-Mini-D2D satellites feature inter-satellite links, so that is a point in favor of Starlink. BlueWalker will be succeeded by five commercial satellites called BlueBirds which are currently under construction.
As for technology, I can't find any information about the Starlink direct-to-cell satellites. As far as I can tell they haven't been even given NORAD IDs yet.The entire Jan 2 Group 7-9 launch seems to be classified as 58699/Object-A. Given that the launch lifted 21 rather than the standard 22 satellites we can assume the new design is similar to the Starlink v2-Mini in size and weight - perhaps only 125 kg heavier. This suggest the efficiency/weight ratio favors Starlink. Starlink v2-Mini weighs 800 kg
edit: better sources:
starlink v2-mini:
* antenna array: 11.07 m2
* solar array: 104.96 m2
* magnitude: +7.87 +- 0.09 (mean, distance-adjusted)
* weight: ~800 kg
starlink v2-mini d2d:
* antenna array: 19.98 m2 (1?)
* solar array: ?
* magnitude: ?
* weight: ~970 kg
* mobile bandwidth: "~7Mb / beam" (40-100km radius)
bluewalker 3:
* antenna array: 64 m2 (front side)
* solar array: 64 m2 (back side)
* magnitude: +3.09 +- 0.04 (mean, distance-adjusted)
* weight: ~1500 kg (redacted?)
* mobile bandwidth: "14 Mbps" (2)
1. I don't have access to the FCC filings so I multiplied the new bus length by the old width.
2. Per-beam, I assume.
From what I gather satellite mass is more correlated with coverage (number of beams) rather than beam efficiency. For the moment, ASTS antennas are just better ($/bandwidth not published). And the new BlueBird satellites are going to be horribly massive and horribly bright compared to BlueWalker-3.
P.S. Anyone know what BlueWalker-3-deb represents?
Discussion here: https://news.ycombinator.com/item?id=38856845