Apple Hires Two Google Executives With Satellite Expertise

[sudo1996]

satellite internet please. get rid of carriers and get rid of the complexity of using my iphone in a new country.

I have it, and it sucks, >9000ms ping. They could improve it, but it's easier to improve cellular.

Edit: I don't actually mean over 9000 ping. It's more like 1000-2000.
[doublepost=1492833998][/doublepost]

exactly this. also, imagine the number of hops would be dramatically reduced once everyone starts getting on satellites. bottleneck would be scalability by the satellites.

... which is pretty bad. I can't imagine it's cheap to upgrade equipment in space.

 

[cube]

Hooray, more space debris.

The UN should regulate space before the cowboys ruin it.

 

[Winni]

I have it, and it sucks, >9000ms ping.

Sounds wrong. 600ms is what you'd normally expect from a satellite link. But then again, with 600ms we're not actually talking about a phone anymore, but about a regular satellite modem and antenna.
[doublepost=1492878625][/doublepost]

... which is pretty bad. I can't imagine it's cheap to upgrade equipment in space.

Yes. You're talking about nine digit numbers per bird.
[doublepost=1492878735][/doublepost]

The UN should regulate space before the cowboys ruin it.

I've worked for the UN. With their level of bureaucracy, NOTHING productive would ever get done anymore - you would have successfully killed any space program.

 

[cube]

I've worked for the UN. With their level of bureaucracy, NOTHING productive would ever get done anymore - you would have successfully killed any space program.

We have the Paris Agreement, no?

 

[scrumpyjack]

satellite internet please. get rid of carriers and get rid of the complexity of using my iphone in a new country.

Wow I'd never even considered this. It'd be an absolute game changer. Roaming fees and swapping sims shouldn't be something we have to endure in 2017. Make it happen Apple.

 

[truthertech]

What ever happened to their anti-poaching pact?

Or was it only when engineers and executives were leaving Apple?

You are about years behind the news. That was an early, informal and illegal restraint of trade, agreement that was ended. It artificially held down wages and Google, Apple and others had to agree to pay several hundred million dollars to engineers, etc., to settle a class action lawsuit.

 

[AppleScruff1]

Would love it if Apple became an ISP. We could cut out all of these companies that are monopolizing and manipulating our laws so they can sell our privacy without our permission. Apple would make a great set of dumb pipes, but it's probably going to cost us. I'd be willing to pay as anything they devise would likely be fast and secure. Satellites though? That sounds like a big challenge.

And then Apple can monopolize. Brilliant.

 

[Tech198]

My guess it's related to car communication.

Hopefully GPS only...

There is no reason for personal information or anything else to be sent over satellites.

 

[gatortpk]

5G (Fifth Generation Wireless Systems) is a mere 3 years away and will easily trump anything 1000-5000 satellites in LEO can do. Though coverage won't be as global, but its rare to need a connection in the middle of nowhere. Fiber lines with 28GHz, 37GHz, and 39GHz bands (All iPv6 with direct device-to-device communication as needed) is the way for the next decade. I don't know what 6G will bring in the '30s but, 5G can connect virtually all devices mobile or not (in the billions all at 1-40 gigabit speeds).

 

[sudo1996]

Sounds wrong. 600ms is what you'd normally expect from a satellite link. But then again, with 600ms we're not actually talking about a phone anymore, but about a regular satellite modem and antenna.

I was exaggerating and making the "over 9000" reference. In my experience, I've had 1000ms-2000ms ping with a satellite modem, but that's only in the one place I've used it, the Mojave Desert. The bandwidth was enough to watch a movie, but the latency made SSH sessions annoying to use.
[doublepost=1492933237][/doublepost]

5G (Fifth Generation Wireless Systems) is a mere 3 years away and will easily trump anything 1000-5000 satellites in LEO can do. Though coverage won't be as global, but its rare to need a connection in the middle of nowhere. Fiber lines with 28GHz, 37GHz, and 39GHz bands (All iPv6 with direct device-to-device communication as needed) is the way for the next decade. I don't know what 6G will bring in the '30s but, 5G can connect virtually all devices mobile or not (in the billions all at 1-40 gigabit speeds).

It's amazing. 4G is already better in practice than wifi in many places, especially when it comes to handoff. Cell tower miles away vs wifi station right in my house, and the wifi doesn't work as reliably.

I do wonder about IPv6. It's really only that useful for countries like China that can't get many v4 addresses, though they could abuse NAT to get more. Everyone messed up the layering system and baked in v4 everywhere in software and such, and many U.S. and Europe companies don't have any need for v6. In fact, those with big "real estate" of v4 addresses are very happy (looking at universities with their /8 subnets).

First thing I do when I set up a PC is disable IPv6 because something along my path to Google's IPv6 address isn't working right, either the router or Comcast's IPv6 routing. I'd often have problems connecting.
[doublepost=1492933471][/doublepost]

I've worked for the UN. With their level of bureaucracy, NOTHING productive would ever get done anymore - you would have successfully killed any space program.

Hey, I've heard that they strongly condemn ISIS.

 

[Tjosansa]

Get rid of celluar isp we have today. Same price all over the globe. Netneutrality. We have a Winner.

 

[Soccertess]

Wish they hired a new CEO too!

Tim can't ride off Steve's coattails forever!

 

[prasand]

Hopefully GPS only...

There is no reason for personal information or anything else to be sent over satellites.

I think it's Car related also, for a wider area communication of shared car data network; non-personal, road data, updated conditions. Like maybe a road exploded somewhere and other cars should learn to navigate the crater. Data can be pushed from car to car within proximity, but what about the rest where cellular data might be prohibitive or wifi isn't available? Or the satellites could be used to provide Car experience, exclusive Apple services for members.

 

[farewelwilliams]

I have it, and it sucks, >9000ms ping. They could improve it, but it's easier to improve cellular.

Edit: I don't actually mean over 9000 ping. It's more like 1000-2000.
[doublepost=1492833998][/doublepost]
... which is pretty bad. I can't imagine it's cheap to upgrade equipment in space.

that's pretty much one of the major goals of SpaceX: to bring down costs of launching stuff into space by re-using rockets instead of throwing them away after one use.

i could easily imagine the cost of launching a rocket being eventually cheaper than deploying/fixing cables in Los Angeles.

 

[sudo1996]

that's pretty much one of the major goals of SpaceX: to bring down costs of launching stuff into space by re-using rockets instead of throwing them away after one use.

i could easily imagine the cost of launching a rocket being eventually cheaper than deploying/fixing cables in Los Angeles.

You don't need to run cables everywhere, just enough to build cell towers. I'm pretty sure the limiting factor for cell in urban areas in terms of speed/cost isn't the cables or towers anymore but the boxes in the stations and towers, and there's new "personal cell" technology on the horizon that claims to address at least the problems with cell towers. As it is now, landline users are often stuck with monopolized cable services, but everyone gets coverage from all cell carriers.
[doublepost=1493088506][/doublepost]

Would love it if Apple became an ISP. We could cut out all of these companies that are monopolizing and manipulating our laws so they can sell our privacy without our permission. Apple would make a great set of dumb pipes, but it's probably going to cost us. I'd be willing to pay as anything they devise would likely be fast and secure. Satellites though? That sounds like a big challenge.

The reason there are monopolies is they are "natural monopolies." One or two ISPs build their lines in an area, and everyone is stuck with them.

 

[macduke]

You don't need to run cables everywhere, just enough to build cell towers. I'm pretty sure the limiting factor for cell in urban areas in terms of speed/cost isn't the cables or towers anymore but the boxes in the stations and towers, and there's new "personal cell" technology on the horizon that claims to address at least the problems with cell towers. As it is now, landline users are often stuck with monopolized cable services, but everyone gets coverage from all cell carriers.
[doublepost=1493088506][/doublepost]
The reason there are monopolies is they are "natural monopolies." One or two ISPs build their lines in an area, and everyone is stuck with them.

No, they actively lobby and get local laws passed to restrict competitors from putting in lines in areas they already have lines. There is nothing natural about that. Google has had to fight this in many areas with their fiber service. Sure, there are some areas that have natural monopolies, but in many larger cities companies would love to compete but are prevented from doing so.

 

[farewelwilliams]

You don't need to run cables everywhere, just enough to build cell towers. I'm pretty sure the limiting factor for cell in urban areas in terms of speed/cost isn't the cables or towers anymore but the boxes in the stations and towers, and there's new "personal cell" technology on the horizon that claims to address at least the problems with cell towers. As it is now, landline users are often stuck with monopolized cable services, but everyone gets coverage from all cell carriers.

current cell towers have extremely bad latency. maybe 5G will solve it client-to-tower latency issues, but you still have the tens/hundreds of hops to communicate with a server across the country. imagining instead with the future satellite internet: you'll probably hop once or twice between satellites to communicate across the country which will more than make up for the latency between the device and satellite.

also you'll be deploying cell towers every 20-40 miles. cost of maintaining that could exceed launching rockets (in the future of course when SpaceX advances).

and speeds of the current towers aren't faster than the current ISP gigabit internet speeds we're getting.

p-cell only serves one customer at a time. it's an interesting technology, but you'll need to deploy tens of thousands of p-cells near a football stadium. and i haven't read anything about the radius it covers so if it's a short distance, you'll need to deploy millions of these things and I don't think carriers are going to shell out cash for that.

 

[sudo1996]

current cell towers have extremely bad latency. maybe 5G will solve it client-to-tower latency issues, but you still have the tens/hundreds of hops to communicate with a server across the country. imagining instead with the future satellite internet: you'll probably hop once or twice between satellites to communicate across the country which will more than make up for the latency between the device and satellite.

A single hop to a satellite from the ground alone is 120ms, which is already longer than the RTT on my phone to/from Google.com. Regardless, it's one thing to implement the network edge with satellites, which is what is currently done, and a totally different challenge to build an entirely satellite Internet, which is what you're suggesting. If it's just the edge, you have the 10 or so hops to the target server regardless. Otherwise, yes you have only 2 or so hops, but I seriously doubt satellites can handle the bandwidth of the Internet backbone you're asking to replace.

I can't find info on the latency of the 4G wireless connection alone, but AFAIK most latency is due to queuing delays in routers, switches, and middleboxes. The best antidote to that is hardware and software that's easy to upgrade over time, and I think putting it all in space makes things a lot harder, no matter how cheap you can make the rockets.

 

[aajeevlin]

A single hop to a satellite from the ground alone is 120ms, which is already longer than the RTT on my phone to/from Google.com. Regardless, it's one thing to implement the network edge with satellites, which is what is currently done, and a totally different challenge to build an entirely satellite Internet, which is what you're suggesting. If it's just the edge, you have the 10 or so hops to the target server regardless. Otherwise, yes you have only 2 or so hops, but I seriously doubt satellites can handle the bandwidth of the Internet backbone you're asking to replace.

Also, AFAIK most latency is due to queuing delays in routers, switches, and middleboxes. The best antidote to that is hardware and software that's easy to upgrade over time, and I think putting it all in space makes things a lot harder, no matter how cheap you can make the rockets.

120ms, are you referring to geostationary satellite? Looking at your math, I think you are. I think we are talking about low Earth orbit here, so altitude around 1200km. With that altitude you can probably have much lower round trip time.

 

[sudo1996]

120ms, are you referring to geostationary satellite? Looking at your math, I think you are. I think we are talking about low Earth orbit here, so altitude around 1200km. With that altitude you can probably have much lower round trip time.

Yes, I was referring to geostationary. You're right about the low orbit one having less latency, and I just saw this: https://arstechnica.com/information...lite-internet-with-low-latency-gigabit-speed/

 

[farewelwilliams]

A single hop to a satellite from the ground alone is 120ms, which is already longer than the RTT on my phone to/from Google.com. Regardless, it's one thing to implement the network edge with satellites, which is what is currently done, and a totally different challenge to build an entirely satellite Internet, which is what you're suggesting. If it's just the edge, you have the 10 or so hops to the target server regardless. Otherwise, yes you have only 2 or so hops, but I seriously doubt satellites can handle the bandwidth of the Internet backbone you're asking to replace.

I can't find info on the latency of the 4G wireless connection alone, but AFAIK most latency is due to queuing delays in routers, switches, and middleboxes. The best antidote to that is hardware and software that's easy to upgrade over time, and I think putting it all in space makes things a lot harder, no matter how cheap you can make the rockets.

SpaceX is planning low orbit satellites with 25 to 35ms latency. I don't know where you live, but perhaps a google cdn is near you? Last I saw, the average latencies of 4G were 100-200ms. and we're talking a couple of hundred thousands of cell towers in the united states that need maintenance.

again, space flight is expensive now, but over time SpaceX will get to the point where all you need to do is pay for fuel to launch something into space by re-using their existing rockets rather than building a new rocket for each launch. we're still years away before rockets will be this affordable, but it certainly isn't impossible.

 

[freediverx]

satellite internet please. get rid of carriers and get rid of the complexity of using my iphone in a new country.

Satellite generally sucks for internet access. Even if they manage to get broadband speeds, the latency is terrible, with current satellite ISPs having latencies of 600ms or more. But with low orbit satellites, that could be reduced to a perfectly reasonable 25-35ms, and it could deliver services to remote areas not served by existing services.

I wonder, though, if a satellite network could maintain broadband speeds while serving hundreds of millions of users simultaneously, and whether one could be deployed at a cost that would make such a service affordable.
[doublepost=1495620996][/doublepost]

What ever happened to their anti-poaching pact?

Or was it only when engineers and executives were leaving Apple?

Apple, Google, others settle antipoaching lawsuit for $415 million
https://www.cnet.com/news/apple-google-others-settle-anti-poaching-lawsuit-for-415-million/

[doublepost=1495621188][/doublepost]

It won't get better. There is a literal bandwidth limit to satellite communications.

There's no theoretical limit to satellite bandwidth. The main limitation is latency, and that can be eliminated with low orbit satellites.

[doublepost=1495621482][/doublepost]

The Terra Bella deal also includes a group of Google's SkySat Earth imaging satellites.

I thought the best aerial imagery was from planes and drones, not sattelites. Commercially available satellite imagery is not very good for detailed views.

[doublepost=1495621814][/doublepost]

Last I saw, the average latencies of 4G were 100-200ms. and we're talking a couple of hundred thousands of cell towers in the united states that need maintenance.

Satellite has great potential to supplement, but not replace cellular. For satellite internet access, you need to be outdoors with a line of sight to the nearest satellite. So it won't work inside your house, your office, any retail store, or even when outdoors in densely packed urban areas.

At best, we will have a hybrid ground/space solution, as we have now for geolocation, with wifi/cellular triangulation providing quick approximate location data that works indoors, and satellites providing a highly precise location fix when you step outside.

 

[freediverx]

5G (Fifth Generation Wireless Systems) is a mere 3 years away and will easily trump anything 1000-5000 satellites in LEO can do. Though coverage won't be as global, but its rare to need a connection in the middle of nowhere.

Tell that to the millions of people who live in areas poorly served by wireless carriers and landline ISPs, and to anyone traveling outside of densely populated areas. Tell that to everyone sick and tired of dealing with customer-hostile near monopolies like Comcast and AT&T.

 

[satcomer]

As a small live sports radio station to us the largest problem is the 2 second delay using a smartphone system! Also when larger games happen (with a large attendance) you could almost forget it! All those phone connecting to one attenna cause 'drop outs and robot voice on Americas largest network!

So something other then cellular is badly needed!

 

[campyguy]

I thought the best aerial imagery was from planes and drones, not sattelites. Commercially available satellite imagery is not very good for detailed views.

Back in the '90s and '00s, yes. Satellite imagery has advanced hugely in the past several years and it's certainly good enough for what most of us (mobile and web users) need/want/consume.

IMO most areas that are developable already have aerials (from planes) already, often from the '00s when developers were going batsh*t crazy. Google used to "lease" that imagery from the repository entities that house that data (for instance, in the Portland Metro area, it's the pseudo-agency named Metro, which offers subscriptions for their data that often gets updated quarterly - we get a new DVD and mail the replaced DVD back to them); maybe you had noticed when Google started out with their mapping, you'd see "Metro" cited as a source - the cost of those "hits" started adding up.

More importantly, with the economic crash (which really started in '05 - I worked with underwriters, and I started advising my clients to get out of housing then…) - and flyovers were no longer being consigned so mapping was getting progressively more-out-of-date. Take a look at the aerial in Apple Maps for these coordinates - 45.6119,-122.6837 (Jantzen Beach Center) - and you'll see a large construction site; the last flyover in that area was done in 2011 and that's what those maps reflect, and that work was completed in 2012 - they're woefully out of date, and there's dozens of new businesses there now. Then look at Google Maps of the same coordinates and surrounding area - I'd offer that "current data" is far more helpful and informative than "out of date data", even if the current data isn't as precise. I used to have an office on that Island and I had to keep telling my techs to stop using Metro's data as it was out of date and not accurate, and use my survey data.

IMHO Google performed a huge favor for all of us by sending those HD satellites up!

Besides, if I need accurate data I call a surveyor. I only use aerials for preliminary data as complimentary support only for a rough survey. If I need analysis data I'll "rent" SHAPE files from the local surveyor/agency (if they exist) and tie my survey in. Cheers!