We are also working on the new build system that will make it easier to create and maintain custom image flavors and will serve as base for Autonomus Build Environment or ABE for customers who require build and serve VyOS in air gapped environments (for example ships including space ones)
Some interesting info : current supported density is 100 Starlinks per 300 sqkm. Slide 6 from https://drive.google.com/file/d/1UOI7b5flgAjJrPs2HDa64p_ZABckasa_/view
Quote from: HarmonicGF2 on 12/22/2021 04:36 pmSome interesting info : current supported density is 100 Starlinks per 300 sqkm. Slide 6 from https://drive.google.com/file/d/1UOI7b5flgAjJrPs2HDa64p_ZABckasa_/view1 cell with 15 miles diametr has square 380 km2 or 130 users This is almost 2 times less than it was calculated here. https://forum.nasaspaceflight.com/index.php?action=post;quote=2308612;topic=48297.3320This means that the average monthly speed per subscriber is about 6-8 Mbps if 1 beam serve only 1 cell.
Quote from: vsatman on 12/23/2021 11:33 amQuote from: HarmonicGF2 on 12/22/2021 04:36 pmSome interesting info : current supported density is 100 Starlinks per 300 sqkm. Slide 6 from https://drive.google.com/file/d/1UOI7b5flgAjJrPs2HDa64p_ZABckasa_/view1 cell with 15 miles diametr has square 380 km2 or 130 users This is almost 2 times less than it was calculated here. https://forum.nasaspaceflight.com/index.php?action=post;quote=2308612;topic=48297.3320This means that the average monthly speed per subscriber is about 6-8 Mbps if 1 beam serve only 1 cell.You do realize that that number is pretty much comparable to municipal broadband (cable, *dsl, and probably even fiber)? High oversubscription is something all ISPs do.In my student dorm we had only 1Mbit/user (700 users, 700mbit/s), but I always achieved download speeds of 100mbit/s if the source allowed it.
I don't think vsatman is saying the speed is low, I think he's saying the opposite.
Quote from: HarmonicGF2 on 12/22/2021 04:36 pmSome interesting info : current supported density is 100 Starlinks per 300 sqkm. Slide 6 from https://drive.google.com/file/d/1UOI7b5flgAjJrPs2HDa64p_ZABckasa_/viewSome very interesting stuff. They plan on 200k terminals in India alone by the end of 2022. Considering that's only one territory, they'll be moving into billion dollar revenue by 2023 at the latest. They also seem confident in being able to supply user terminals in very large quantities.
Quote from: niwax on 12/22/2021 05:45 pmQuote from: HarmonicGF2 on 12/22/2021 04:36 pmSome interesting info : current supported density is 100 Starlinks per 300 sqkm. Slide 6 from https://drive.google.com/file/d/1UOI7b5flgAjJrPs2HDa64p_ZABckasa_/viewSome very interesting stuff. They plan on 200k terminals in India alone by the end of 2022. Considering that's only one territory, they'll be moving into billion dollar revenue by 2023 at the latest. They also seem confident in being able to supply user terminals in very large quantities.I would hesitate to extrapolate cleanly from North America/Europe prices to Indian prices. Many companies offer services much closer to cost there and in other lower income markets (I know this from personal experience). It’s possible Starlink is charging the same price - there are quite a few well off Indians and Indian businesses - but I’d want it confirmed if possible.
I would hesitate to extrapolate cleanly from North America/Europe prices to Indian prices. Many companies offer services much closer to cost there and in other lower income markets (I know this from personal experience). It’s possible Starlink is charging the same price - there are quite a few well off Indians and Indian businesses - but I’d want it confirmed if possible.
Although it's also possible that Starlink is still giving each subscriber 2~3 Mbit/s but is using one beam to serve multiple cells.
Quote from: su27k on 12/24/2021 04:33 amAlthough it's also possible that Starlink is still giving each subscriber 2~3 Mbit/s but is using one beam to serve multiple cells.This is a good point. Due to the limited number of satellites in service at present, it seems almost certain that they are serving multiple cells with each beam (or at least with each sat-side antenna, which is slightly different since 1 antenna can potentially form multiple beams). For example, in the US, they have about 50 satellites overhead at any given time, each with 4 antennas. But the US is 8 million square km, or over 25,000 cells in area. So they would need to serve 133 cells with each antenna to offer complete coverage, or at one antenna per cell could only cover less than 1% of the country.Depending on how much angular separation they need to reuse the same channel, though, they might be able to get a lot more than 600-800 Mbps per antenna/channel.
2) But more importantly, you have only 2000 MHz frequency band in each polarization on thelink Gateway -satellite in the Ka band, this band is converted on the satellite to the Ku band on the link satellite- user terminal - 8 beams of 240 MHz in one right polarization
the beam can continuously serve cell N1 for 10 milliseconds, then switch to the cell N2 (switching time for FAR ESA is about only 10 microseconds) to transmitt 10 milliseconds to UT in Cell N2 and switch back to cell N 1. N1 - N2 -N1 - N2 .....and then the Sputnik will serve not 8 cells, but 16.But the average speed and volume of transmitted traffic will also be 2 times less for each of the 16 cells than in the version with 8 cells. But if you have not so many subscribers in each cell, not 200-300, but 50-100, then this approach will help you serve more subscribers and close more territory.theoretically, the beam can serve not 2 cells, but more, for example, as in the figure below 8, but then you need to reduce the operating time in each cell to 2-5 milliseconds, otherwise there will be a very high latency
1) They have limited spectrum for the uplink, but (unlike at the user terminal receiver) there are no arbitrary limits on flux density or aperture size at the satellite receiver. So the SNR can be arbitrarily high as needed, and the power flux and aperture can be sized to make the uplink data throughput match the downlink. It doesn't seem to me that uplink spectrum is a limiting issue at the moment.2)Or, since each antenna array is really thousands of small antennas, they could subdivide the arrays into 2 or more smaller or sparser arrays and transmit multiple beams simultaneously in different directions. This assumes they have the necessary margins on power and beam angle, since the small or sparse array will perform worse on one or more of those factors.
Elon Musk has hit back at criticism that his company’s Starlink satellites are hogging too much room in space, and has instead argued there could be room for “tens of billions” of spacecraft in orbits close to Earth.“Space is just extremely enormous, and satellites are very tiny,” Musk said. “This is not some situation where we’re effectively blocking others in any way. We’ve not blocked anyone from doing anything, nor do we expect to.”His comments, made in an interview with the Financial Times, came in response to a claim from Josef Aschbacher, head of the European Space Agency, that Musk was “making the rules” for the new commercial space economy. Speaking to the FT earlier this month, Aschbacher warned Musk’s rush to launch thousands of communications satellites would leave fewer radio frequencies and orbital slots available for everyone else.
