Apple maybe getting interested in broadband satellites

[Star One]

Surely pretty inevitable considering the interest of other large technology companies in this area.

Apple said to have held talks about investing in Boeing satellite broadband project

https://www.bloomberg.com/news/articles/2017-04-21/apple-hires-top-google-satellite-executives-for-new-hardware-team

[Eric Hedman]

This is getting interesting:

https://www.bloomberg.com/news/articles/2019-12-20/apple-has-top-secret-team-working-on-internet-satellites

https://www.dailymail.co.uk/sciencetech/article-7814721/Apple-secret-plan-develop-satellites-beam-data-space-report-says.html

[Eric Hedman]

I've had a little while to think about this.  Plus I a few minutes ago I drove past a Verizon store.  This would allow Apple to bypass the cell phone companies and get the whole pie.  If I were Verizon, AT&T, Sprint, etc. I would be getting very nervous.  Their biggest partner might now become a competitor.  Samsung and other cell phone makers are probably having emergency meetings right now if they didn't already know about this.  It may actually be a lot cheaper to build a constellation than to maintain cell towers everywhere.  I also wonder launch provider(s) they would go with.

Disruptions to industry are fun to watch as long as you aren't the one being disrupted.  It will be fascinating to see if this turns into a major industry disruption in the next few years.

[RedLineTrain]

The descriptions in the articles are vague, but if they are aiming to replace cell towers, this ground has been tread before.  The most notable is Iridium.

If Apple is serious on replacing towers, they should just buy Iridium (that has the necessary spectrum rights worldwide) and then embark on a Super-Iridium constellation.

But I doubt they are serious, or serious enough, for this market.  It reminds me of their supposed entry into the electric car business.

[Coastal Ron]

From the Bloomberg article:

It’s not clear if Apple intends to pursue the costly development of a satellite constellation itself or simply harness on-the-ground equipment that would take data from existing satellites and send it to mobile devices.

As I recall it is possible for current generation cellphones to detect satellite signals, but the software that manages detecting RF signals ignores signals that are not local - in other words, the handshake connections time out because of the additional distance there is between the surface of the Earth and even LEO.

Modern cellphones already use assisted GPS, which is a hybrid of GPS satellites and terrestrial networks, and what I'm describing is a different data network separate from GPS.

Note, I'm not an engineer, and I'm going off of memory of an article that I read more than a year ago, but I remembered it because of the implications that standard cellphones could get more than GPS data from satellites - if their software were changed to allow it.

I think that could be one possibility that Apple is looking at, and it would be something that could be deployed through regular model upgrades, and without the need to own their own satellite network.

/wild*speculation

[Exastro]

More speculation:


The plan is to build a constellation of ~100 satellites in MEO and roll out service around 2022.  Initial service will be to high-end iPhones only.  These iPhones will have optimizations that significantly improve service quality over generic phones, and their limited number will keep the constellation from being inundated early on.  MEO lets them cover the globe without having to launch ~1000 satellites while still keeping the latency and antenna mass acceptably low.

The service will be low-rate data and voice only.  It'll differentiate itself from Iridium by smaller, sleeker, handsets that double as ordinary smartphones, and little emphasis will be made on IoT applications.  It'll be marketed as an obviously upscale product that lets the user transcend national boundaries and get a signal basically anywhere he can see the sky.  The revenues generated at this stage would be modest, and probably not enough to justify the investment.

If the rollout goes well, the plan is to license the critical parts of the terminal design to other smartphone makers and launch additional satellites to support the growing traffic.  This would give Apple a potentially valuable pair of revenue streams (connection fees and license fees).  It would also give them first-mover advantage if the expanded offerings sell well, which means they'd have a lot of influence in the further development of the basic technology.

[gongora]

I only read about half of the Bloomberg article, didn't think it was particularly well written/researched.

Completely random speculation is unlikely to be useful.

Another company is already actively testing connectivity to phones from orbit:
https://lynk.world/  https://forum.nasaspaceflight.com/index.php?topic=47511.0

[arachnitect]

The imagery people involved make me suspect this is more about mapping/traffic than data connections with handsets.

I think Apple is making a push on 3D mapping to better compete with Google, position for Augmented Reality, and further down the line Autonomous Vehicles.

[TrevorMonty]

Apple has spare cash ($245B) to test concept on small constellation.

[freddo411]

I've had a little while to think about this.  Plus I a few minutes ago I drove past a Verizon store.  This would allow Apple to bypass the cell phone companies and get the whole pie.  If I were Verizon, AT&T, Sprint, etc. I would be getting very nervous.  Their biggest partner might now become a competitor.  Samsung and other cell phone makers are probably having emergency meetings right now if they didn't already know about this.  It may actually be a lot cheaper to build a constellation than to maintain cell towers everywhere.  I also wonder launch provider(s) they would go with.

Disruptions to industry are fun to watch as long as you aren't the one being disrupted.  It will be fascinating to see if this turns into a major industry disruption in the next few years.

I think we are quite far technically from enabling large scale mobile phone service solely from satellites.   Terrestrial towers will be a very important part of any cell service in urban areas.

That being said, rural service may be dominated by satellites much quicker than people imagine.

[su27k]

But I doubt they are serious, or serious enough, for this market.  It reminds me of their supposed entry into the electric car business.

Exactly, their adventures into cars is going nowhere.

And I'm not sure if it's even feasible for satellites to replace cell towers, how does that work when you're inside buildings?

[Eric Hedman]

But I doubt they are serious, or serious enough, for this market.  It reminds me of their supposed entry into the electric car business.

Exactly, their adventures into cars is going nowhere.

And I'm not sure if it's even feasible for satellites to replace cell towers, how does that work when you're inside buildings?

Apple has jumped into quite a few venture with either starting something new or by acquisition.  Some of them have been quite successful.  Others just faded away.  The list of the acquisitions is here:

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

Just because they gave up on cars doesn't mean anything when it comes to what they are looking into now.

"And I'm not sure if it's even feasible for satellites to replace cell towers, how does that work when you're inside buildings?" It works just like cell towers do if you're inside a building.  If you are on the right frequency and your broadcast power is sufficient and you have a sensitive enough receiver it works.  In some buildings cell phones don't work for various reasons.  I've been in a few where Sprint put in repeaters.  There are several antennas hanging inside and a large one on the roof to connect to cell towers.  The same could be done with buildings that block enough of the signal with satellites.

[Nomadd]

But I doubt they are serious, or serious enough, for this market.  It reminds me of their supposed entry into the electric car business.

Exactly, their adventures into cars is going nowhere.

And I'm not sure if it's even feasible for satellites to replace cell towers, how does that work when you're inside buildings?

Small, local APs, similar to the femto/micro cells phone companies send out to customers.

[noogie]

There are too many issues with the physics issues for them to make using their own satellites a feasible alternative to cellular networks.
I think it will be for other things like mapping (they need to improve Apple Maps) as well as remote sensing and geospatial data for augmented reality (which they have been interesting in adding to the iphone for a while now).
Of course it could just be some research that may never see the light of day (something that may come under "other bets" at Google but at Apple stays secret until leaked like this). Of course for Apple the cost of having a team like this go nowhere is chump change 

[TrevorMonty]

There are too many issues with the physics issues for them to make using their own satellites a feasible alternative to cellular networks.
I think it will be for other things like mapping (they need to improve Apple Maps) as well as remote sensing and geospatial data for augmented reality (which they have been interesting in adding to the iphone for a while now).
Of course it could just be some research that may never see the light of day (something that may come under "other bets" at Google but at Apple stays secret until leaked like this). Of course for Apple the cost of having a team like this go nowhere is chump change 

These big companies spend a lot of money on researching brainstorming ideas like this. Most go nowhere but occasionally they hit jackpot eg Amazon AWS. Only need one AWS to justify a 100 projects that go nowhere.

[Robotbeat]

But I doubt they are serious, or serious enough, for this market.  It reminds me of their supposed entry into the electric car business.

Exactly, their adventures into cars is going nowhere.

And I'm not sure if it's even feasible for satellites to replace cell towers, how does that work when you're inside buildings?

This is a lot closer to their core business.

[Asteroza]

I only read about half of the Bloomberg article, didn't think it was particularly well written/researched.

Completely random speculation is unlikely to be useful.

Another company is already actively testing connectivity to phones from orbit:
https://lynk.world/  https://forum.nasaspaceflight.com/index.php?topic=47511.0

Ubiquitilink/Lynk Global has the operational experience with direct cellular with a satellite via Cygnus experiments, so they are are the farthest along.

It would not be uncharacteristic of Apple to underwrite a "supplier" completely to stand up new capability for the Apple ecosystem. See what happened with the artificial ruby factory for when that goes wrong, but Apple has invested in factories for supplies in exchange for manufacturing exclusivity with considerable success, gaining access to new parts well before competitors. Apple going long and setting up a 5G exclusive satellite network would be an interesting play to bypass the capital tables problem incumbent cellular carriers face in recapitalizing for 5G deployment. Apple already has their eSIM MVNO service riding on top of other carriers as is, so an interesting end run would be offloading wherever the MVNO can operate, but where traffic allows, push to satellite.

One major issue is setting up the large antennas on such satellites for such such a large constellation of low flying sats. The only realistic solutions are things like Archinaut, or SpiderFab (particularly the open documentation regarding Tethers Unlimited concepts for OrbWeaver, which specifically envisioned extremely large antennas on sats). Getting ITU clearance to run pure 5G direct from satellites will be an interesting legal adventure however.

Apple doing a double grab of TUI and Lynk.Global to get the ball rolling would be very interesting...

[ChrisWilson68]

But I doubt they are serious, or serious enough, for this market.  It reminds me of their supposed entry into the electric car business.

Exactly, their adventures into cars is going nowhere.

And I'm not sure if it's even feasible for satellites to replace cell towers, how does that work when you're inside buildings?

This is a lot closer to their core business.

I agree that it's closer to their core business, but it's still pretty far outside their core competency.  Designing and building a satellite system is not the same as building a consumer product.

[Asteroza]

Designing and building a satellite system is not the same as building a consumer product.

Nobody tell Amazon that...

[TrevorMonty]

I only read about half of the Bloomberg article, didn't think it was particularly well written/researched.

Completely random speculation is unlikely to be useful.

Another company is already actively testing connectivity to phones from orbit:
https://lynk.world/  https://forum.nasaspaceflight.com/index.php?topic=47511.0

Ubiquitilink/Lynk Global has the operational experience with direct cellular with a satellite via Cygnus experiments, so they are are the farthest along.

It would not be uncharacteristic of Apple to underwrite a "supplier" completely to stand up new capability for the Apple ecosystem. See what happened with the artificial ruby factory for when that goes wrong, but Apple has invested in factories for supplies in exchange for manufacturing exclusivity with considerable success, gaining access to new parts well before competitors. Apple going long and setting up a 5G exclusive satellite network would be an interesting play to bypass the capital tables problem incumbent cellular carriers face in recapitalizing for 5G deployment. Apple already has their eSIM MVNO service riding on top of other carriers as is, so an interesting end run would be offloading wherever the MVNO can operate, but where traffic allows, push to satellite.

One major issue is setting up the large antennas on such satellites for such such a large constellation of low flying sats. The only realistic solutions are things like Archinaut, or SpiderFab (particularly the open documentation regarding Tethers Unlimited concepts for OrbWeaver, which specifically envisioned extremely large antennas on sats). Getting ITU clearance to run pure 5G direct from satellites will be an interesting legal adventure however.

Apple doing a double grab of TUI and Lynk.Global to get the ball rolling would be very interesting...

But I doubt they are serious, or serious enough, for this market.  It reminds me of their supposed entry into the electric car business.

Exactly, their adventures into cars is going nowhere.

And I'm not sure if it's even feasible for satellites to replace cell towers, how does that work when you're inside buildings?

This is a lot closer to their core business.

I agree that it's closer to their core business, but it's still pretty far outside their core competency.  Designing and building a satellite system is not the same as building a consumer product.

Don't need to reinvent wheel. Apple can buy a satellite manufacturer or two including operators and their ground system.

[ChrisWilson68]

Designing and building a satellite system is not the same as building a consumer product.

Nobody tell Amazon that...

Amazon has consumer products, but they also run AWS and their own enormous logistics system.  I'd say that's a lot closer to a satellite internet system than anything Apple has done.

[RocketGoBoom]

It is my understanding that the first two constellations to establish themselves in LEO get the first choice of prime spectrum in the Ku and Ka band. From all reports, that seems most likely to be SpaceX Starlink and OneWeb which grab the first two spots.

Anyone that gets their LEO constellation into orbit after that would need to design their system so as not to interfere with the existing constellations using that Ku and Ka spectrum.

So now along comes Amazon Kuiper and wants to skip all of that.
https://www.geekwire.com/2019/spacex-seeks-block-amazons-project-kuiper-satellite-broadband-program-skipping-regulatory-hurdle/

Amazon's play seems like a long shot move and will likely be denied, especially considering that Amazon is likely 3 to 4 years away from launching their first satellites.

So how would players late to the game even begin to compete? Whenever there is a conflict for spectrum, the first movers (Starlink and OneWeb) get the prime spectrum.

That leads me to believe that we will likely see Apple or Amazon probably try to buy OneWeb eventually, if they are really serious about putting a constellation into LEO.

OneWeb is owned mostly by SoftBank and based on recent news, SoftBank is having capital problems funding their unicorn startups. WeWork's collapsed IPO has put SoftBank into a cash critical situation.

The bottom line is, don't be surprised if SoftBank is willing to sell OneWeb (and it's likely top 2 prime spectrum position). That is extremely valuable spectrum which Apple or Amazon could easily afford to purchase if they really decide they want to compete with Starlink in this market.

[Asteroza]

It is my understanding that the first two constellations to establish themselves in LEO get the first choice of prime spectrum in the Ku and Ka band. From all reports, that seems most likely to be SpaceX Starlink and OneWeb which grab the first two spots.

Anyone that gets their LEO constellation into orbit after that would need to design their system so as not to interfere with the existing constellations using that Ku and Ka spectrum.

So now along comes Amazon Kuiper and wants to skip all of that.
https://www.geekwire.com/2019/spacex-seeks-block-amazons-project-kuiper-satellite-broadband-program-skipping-regulatory-hurdle/

Amazon's play seems like a long shot move and will likely be denied, especially considering that Amazon is likely 3 to 4 years away from launching their first satellites.

So how would players late to the game even begin to compete? Whenever there is a conflict for spectrum, the first movers (Starlink and OneWeb) get the prime spectrum.

That leads me to believe that we will likely see Apple or Amazon probably try to buy OneWeb eventually, if they are really serious about putting a constellation into LEO.

OneWeb is owned mostly by SoftBank and based on recent news, SoftBank is having capital problems funding their unicorn startups. WeWork's collapsed IPO has put SoftBank into a cash critical situation.

The bottom line is, don't be surprised if SoftBank is willing to sell OneWeb (and it's likely top 2 prime spectrum position). That is extremely valuable spectrum which Apple or Amazon could easily afford to purchase if they really decide they want to compete with Starlink in this market.

Direct satellite 5G I believe uses different bands from the current megaconstellations, so there would be no benefit from salvaging OneWeb for Apple except as backhaul for terrestrial towers, but Apple would specifically want to avoid terrestrial towers on their balance sheet where possible. Offering a tit-for-tat arrangement to provide backhaul to carriers who allow Apple's eSIM MVNO to roam on them would make some sense though. I doubt you could reconfigure existing sats to do the 5G bands well even with the SDR systems onboard, as the phased array antennas were optimized for specific bands.

There's also the legal barriers, as even though you are deconflicted with the megaconstellations, direct 5G will require a ton of ITU negotiations due to terrestrial interference issues. The incumbent cellular carriers won't take this lying down either, and their legal team backbench are probably at equal fighting strength to Apple's. Apple would somehow have to sweeten the MVNO deal with carriers to get them to back down, but how since this is directly encroaching on their high margin territory.

[gongora]

It is my understanding that the first two constellations to establish themselves in LEO get the first choice of prime spectrum in the Ku and Ka band. From all reports, that seems most likely to be SpaceX Starlink and OneWeb which grab the first two spots.

Anyone that gets their LEO constellation into orbit after that would need to design their system so as not to interfere with the existing constellations using that Ku and Ka spectrum.

This is not correct.  Worldwide the first to file with ITU for each set of frequencies (and subsequently put at least one satellite in orbit) gets some kind of priority, and that is not SpaceX.  OneWeb and Telesat have ITU priority in various frequencies.  SpaceX keeps changing their orbits so they may be at the back of the line as far as ITU filings are concerned.  For coverage of the U.S., everyone in the same processing round has to cooperate on an equal footing.  The first to bring the frequencies into use gets first choice of frequencies when two satellites come near each other and have to split the available bandwidth between them.  (SpaceX was the third to get a satellite into an applicable orbit.  They're trying to argue that they're the first to communicate with FCC filed ground stations in the U.S., which the first two companies, both non-US, say is bulls---.  Legal staffs are bickering over the matter.)

So now along comes Amazon Kuiper and wants to skip all of that.
https://www.geekwire.com/2019/spacex-seeks-block-amazons-project-kuiper-satellite-broadband-program-skipping-regulatory-hurdle/

Amazon's play seems like a long shot move and will likely be denied, especially considering that Amazon is likely 3 to 4 years away from launching their first satellites.

This is correct, and it's not just SpaceX opposing it.  Pretty much everyone from the existing processing round filed opposition to Amazon.  (Amazon's argument is amusing.  They say that their constellation is about as big as the previously filed constellations that withdrew their filings, so the environment with Amazon will just be the same as if those other constellations had been built, right?)

So how would players late to the game even begin to compete? Whenever there is a conflict for spectrum, the first movers (Starlink and OneWeb) get the prime spectrum.

That leads me to believe that we will likely see Apple or Amazon probably try to buy OneWeb eventually, if they are really serious about putting a constellation into LEO.

OneWeb is owned mostly by SoftBank and based on recent news, SoftBank is having capital problems funding their unicorn startups. WeWork's collapsed IPO has put SoftBank into a cash critical situation.

The bottom line is, don't be surprised if SoftBank is willing to sell OneWeb (and it's likely top 2 prime spectrum position). That is extremely valuable spectrum which Apple or Amazon could easily afford to purchase if they really decide they want to compete with Starlink in this market.

How subsequently filed constellations will fit into the picture is a big question right now.  If another processing round is started by the FCC then the new filings would probably be at a disadvantage when communicating with U.S. ground stations, but we'll just have to see how it plays out.  If another round is started you can bet both SpaceX and OneWeb will be filing for more satellites in addition to the new companies.

[RocketGoBoom]

This is not correct.  Worldwide the first to file with ITU for each set of frequencies (and subsequently put at least one satellite in orbit) gets some kind of priority, and that is not SpaceX.  OneWeb and Telesat have ITU priority in various frequencies.  SpaceX keeps changing their orbits so they may be at the back of the line as far as ITU filings are concerned.  For coverage of the U.S., everyone in the same processing round has to cooperate on an equal footing.  The first to bring the frequencies into use gets first choice of frequencies when two satellites come near each other and have to split the available bandwidth between them.  (SpaceX was the third to get a satellite into an applicable orbit.  They're trying to argue that they're the first to communicate with FCC filed ground stations in the U.S., which the first two companies, both non-US, say is bulls---.  Legal staffs are bickering over the matter.)

It is my understanding that the ITU just changed the requirements for everyone last month? Now the requirement is not just one satellite in orbit. It is a certain percentage of the proposed constellation has to be in orbit by certain dates after the application.

11/20/19
https://spacenews.com/itu-sets-milestones-for-megaconstellations/

Regulators meeting in Sharm el-Sheikh, Egypt, for the 2019 World Radiocommunication Conference this month said that for non-geosynchronous constellation operators to keep their full spectrum rights in the future, they will have to hit deployment milestones that start seven years after requesting the spectrum.

After those seven years, NGSO constellation operators will need to launch 10% of their satellites in two years, 50% in five years and 100% in seven years. If constellation ventures fail to launch enough satellites by the milestones, or within the total 14 years allotted, their spectrum rights are limited proportionally to the number launched before time ran out.

Under the new rules, that would favor whoever gets their full constellation into orbit first. The race to reach LEO and squat on your territory seems to be the basic guideline.
 
From the article, it does not yet seem clear who has priority.
Does the first to LEO have priority? Or does the first application have priority, even if they reach orbit WAY later, so long as they meet the deadlines.

I suspect the priority will be given to the first to reach LEO with the required percentage of the constellation. Whoever is launching later would seem to be under the burden to not interfere with the earlier in LEO constellation. The response of Kepler, which was first to orbit with one satellite, seems to underscore this. They have changed their orbital plans due to SpaceX already being in their preferred orbit.

I don't know the answers for sure. But I am reading and trying to understand it all.

[gongora]

First to file still has priority (with ITU) as long as they deploy their satellites.  The deadlines are still pretty long, it didn't say anything about shortening the bringing into use deadline, so you have over a decade to build out the constellation.

[watermod]

But I doubt they are serious, or serious enough, for this market.  It reminds me of their supposed entry into the electric car business.

Exactly, their adventures into cars is going nowhere.

And I'm not sure if it's even feasible for satellites to replace cell towers, how does that work when you're inside buildings?

Small, local APs, similar to the femto/micro cells phone companies send out to customers.

Those femto/micro cells could likely be plugged into the gb-ethernet ports or the wifi on StarLink user devices. 

[Asteroza]

A point in the Bloomberg article that stands out is the founder of Ettus research, which is a premier SDR radio maker, joined the Apple team. There is a whole lot going on the SDR world, so having him is bringing in the big guns. This works both ways too, after Apple bought up Intel's failed 5G radio modem division. We may see interesting innovation in both the satellite end and in the ground terminal end. Combine the SDR work with stuff like diode controlled metamaterial fractal antennas from the likes of Fractenna for new capabilities.