Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

We know a little more about Amazon's satellites

arstechnica.com

Post reply on HN
Print view View original

91–100 of 122 posts

Re: We know a little more about Amazon's satellites

#91

Post by bitmasher9 »

tgsovlerkhgsel wrote: post #8

I wonder what "earth observation" opportunities there would be with such megaconstellations, from simply having a camera with a telephoto lens pointed down to a giant, sky-spanning Synthetic Aperture Radar utilizing multiple satellites. Anything like that would explain the secrecy...

It’s a 450km orbit. Cameras are good, but you’d need quite a bit of mass per satellite to identify anything of interest that isn’t already covered by other satellites. At a certain point photography becomes more a matter of optics (using lens to collect light) than anything else.

Re: We know a little more about Amazon's satellites

#92

Post by mrshadowgoose »

croes wrote: post #90
mrshadowgoose wrote: post #73

Earlier quoted context omitted.

I am entirely convinced that absent LEO comsat constellations, people who espouse this sentiment would likely be whining about "useless astronomy taking money away from helping poor people". If you genuinely care about the field of astronomy, rest assured that the same falling launch costs that have enabled LEO comsat constellations, will enable the launch of fleets of space-based telescopes.

Space based telescopes have limits Earth bound telescopes don’t have and they are easier to maintain

Yes, I am quite aware that the current generation of space-based telescopes are quite limited. And it's solely due to the historically extreme cost of mass to orbit.

The largest proposed ground observatories already use segmented mirrors. One can use the same approach in space, it's only a matter of launch cost.

Re: We know a little more about Amazon's satellites

#93

Post by timewizard »

mrshadowgoose wrote: post #89
timewizard wrote: post #75

Earlier quoted context omitted.

> fleets of space-based telescopes. Isn't one of the nice aspects of astronomy is that you can do quite a bit as an amateur with some decent equipment and a nice vantage point? What value does this fleet have to these people? > people who espouse this sentiment would likely be whining about "useless astronomy taking money away from helping poor people". You've constructed a strawman for the purposes of gatekeeping; m…

> Isn't one of the nice aspects of astronomy is that you can do quite a bit as an amateur with some decent equipment and a nice vantage point? What value does this fleet have to these people? It doesn't, and admittedly I don't really care that much. I care far far more that remote communities can now have meaningful access to the internet, one of the most transformative and enabling technologies in existence, than ni…

> I care far far more that remote communities can now have meaningful access to the internet

Then can you tell me how many remote communities were not being served before that are now suddenly capable of accessing the internet now that these particular constellations exist? I mean just looking at Starlink's current availability map shows how little you might actually care about this particular outcome.

Even so was this the most affordable and sustainable option for these countries? Was there absolutely no way to achieve both goals at once?

> I don't really care that much.

Noted. We're just picking sides today, I guess. Bummer.

Re: We know a little more about Amazon's satellites

#94

Post by bitmasher9 »

fidotron wrote: post #12
greatgib wrote: post #9

[flagged]

And every country as well.

I doubt we’ll end up with 200+ of these consolations in 100 years. Probably not even 20.

It’s just a physics problem. Rocket launches are expensive from an energy standpoint. These satellites will have a decaying orbit that requires replacement. It won’t look appealing to most net-importers of energy (which is most countries, but the whole EU might bear the cost for one network for strategic reasons).

Not to mention most countries just don’t operate enough military assets outside of their borders to justify their own network. Non-military applications will be just fine with E2E encryption over public channels. More localized military operations can have communication needs served other ways.

Re: We know a little more about Amazon's satellites

#95

Post by SEJeff »

neuroelectron wrote: post #70
parsimo2010 wrote: post #41

Earlier quoted context omitted.

They will undoubtedly sell some aspect of Kuiper through AWS. They already have IP addresses and DNS in the AWS product list, and they have all kinds of data transport services. I don’t know if the government implication is as big as you think, as the US government has been doing secure satellite communications for decades and has already given SpaceX the contract for Starshield. So undoubtedly Kuiper would love a pi…

I didn't know about Starshield. I thought Starlink was supposed to be neutral.

Starlink is neutral. Starshield is not. Starshield runs on different satellites with potential for custom additional payloads as well.

Re: We know a little more about Amazon's satellites

#96

Post by saltcured »

lxgr wrote: post #64
nine_k wrote: post #51

Earlier quoted context omitted.

> easily tappable I suppose in any realistic scenario we should assume that the enemy may be listening to all our communication at all times. This is the assumption behind such daily things as WPA3, SSH, TLS. Jamming is a much more serious concern.

Yes, so all in all for satellite vs. fiber in backbone applications, I'd say that it's a wash (or slight win for fiber) when it comes to security, and a definitive win for fiber when it comes to jamming resistance. In the field it's a completely different story, of course – you can't always pull fiber (although it does appear in unexpected scenarios, such as fiber-operated UAVs or torpedoes).

Not sure. Jamming a wireless signal is an active process that stops interfering when you stop transmitting.

Destroying fiber with a backhoe or an axe doesn't stop interfering when you stop digging or chopping though.

Re: We know a little more about Amazon's satellites

#97

Post by SEJeff »

lxgr wrote: post #49
neuroelectron wrote: post #31

The article doesn't even mention AWS, which I think is the obvious implication being overlooked here. AWS has Government and defense contracts, with DoD notably. This enables secure private communication outside of the internet across data centers (of which there are a lot) and of course, to any point on the Earth. The idea that this is for "underserved communities" is probably a sly nod to battlefield logistics.

> This enables secure private communication outside of the internet across data centers How so? I'd imagine the datacenter terminal side downlink to be much more easily tappable than fiberoptics. There are advantages in latency and potentially availability, but even there I would imagine fiber to win in an adversarial active jamming scenario.

FHSS[1] has made jamming difficult in US military communications for decades. It doesn’t make it impossible but jamming the entire spectrum is nearly impossible at scale for almost everyone. At best it would affect small areas until the US sent rf seeking missiles (HARM are designed for this) at the jammer source. Also note that modern satcom like Starlink uses AESA digital phased array antennas much like a F35’s radar. It’s so much more complex than legacy analog stuff.

[1] https://en.m.wikipedia.org/wiki/Frequency-hopping_spread_spe...

Re: We know a little more about Amazon's satellites

#98

Post by SEJeff »

entropie wrote: post #28
perihelions wrote: post #21

Earlier quoted context omitted.

The observation cadence could be game-changing. Instead of once-, twice- daily revisit times, in principle you could contemplate continuous observation, of large parts of the Earth, from LEO, with enough downstream bandwidth to make interesting use of all that data.

Ive seen a video around 2005 about USA spy satellites/drones where they kind of disclosured what was possible at this time. Having a very wide area with realtime object/person tracking and multiple terrabytes of data every minute while beeing able to go back in time with all this features. This all was like 20 years go. 20. 20!! Than I see my upper consumer grade canon camera, a r6mkII with 70-200mm lens (mk1, 20 yea…

They’ve done similar things with hundreds of cell phone cameras and digital techniques to piece it all together.

The prototype was called Gorgon Stare[1] and could surveil an entire city at once.

[1] https://www.sncorp.com/capabilities/wide-area-motion-imagery...

Re: We know a little more about Amazon's satellites

#99

Post by wmf »

tgsovlerkhgsel wrote: post #8

I wonder what "earth observation" opportunities there would be with such megaconstellations, from simply having a camera with a telephoto lens pointed down to a giant, sky-spanning Synthetic Aperture Radar utilizing multiple satellites. Anything like that would explain the secrecy...

Apparently you can't hide from Starshield so I'm guessing it's pretty good. https://arstechnica.com/science/2024/11/nro-chief-you-cant-h...

Re: We know a little more about Amazon's satellites

#100

Post by wmf »

eastbound wrote: post #68
daymanstep wrote: post #22

Earlier quoted context omitted.

I see a minimum price of $325k I don't think a student is going to be able to afford that any time soon

325k$ is for 50kg. Yes it’s the minimum for this program. The students I had were programming the 10cm^3 satellites, I don’t have more details than this cursory explanation but it’s definitely possible.

A cubesat isn't 50kg. Maybe people are subletting a single slot.
Post reply on HN