Quote from: raptorx2 on 03/02/2024 09:49 pmQuote from: abaddon on 03/02/2024 08:29 pmLoss rate is pretty high, this looks like UDP, so I would expect sustained throughput of TCP with retransmissions etc would be substantially lower. QUIC would probably perform better than TCP here. Still pretty cool.LTE Stack development is still in it's infancy. A very steep learning curve. Remember, this is running on low-power constrained processing on the satellite. However, learning Stack development should allow them to deploy other "custom" air interfaces for ultra secure government communications. IMONot exactly "infancy". LTE is deployed on billions of devices and has been around for a decade, with roots going back at least three decades. Quite a few developers are familiar with it, and several stacks are available for purchase.What I don't know is how it is being used for satellite. The lower layers are highly optimized for comms between a cell tower and a cell phone. I was very surprised to learn that the basic lower-layer packet error rate is 10%. I'm used to error rates below 0.1%. The LTE trick is fast NAK and retransmit between the phone and cell tower, and since the distance is so low the recovery is instantaneous. The other trick is that the receiver keeps the bad block. The retransmitted packet uses a different FEC than the first one, and is designed so the two packets together form a longer and more robust FEC block. The problem I have is that this trick won't work nearly as well over a 300 km link, so they will be doing something else.
Quote from: abaddon on 03/02/2024 08:29 pmLoss rate is pretty high, this looks like UDP, so I would expect sustained throughput of TCP with retransmissions etc would be substantially lower. QUIC would probably perform better than TCP here. Still pretty cool.LTE Stack development is still in it's infancy. A very steep learning curve. Remember, this is running on low-power constrained processing on the satellite. However, learning Stack development should allow them to deploy other "custom" air interfaces for ultra secure government communications. IMO
Loss rate is pretty high, this looks like UDP, so I would expect sustained throughput of TCP with retransmissions etc would be substantially lower. QUIC would probably perform better than TCP here. Still pretty cool.
https://twitter.com/elonmusk/status/1764032892663906313QuoteSpaceX just achieved peak download speed of 17Mb/s from satellite direct to unmodified Samsung Android phone
SpaceX just achieved peak download speed of 17Mb/s from satellite direct to unmodified Samsung Android phone
The six test satellites for Starlink's direct-to-cell system were launched as part of Group 7-9 in Jan. The other sats on the launch (blue) have raised to 550 km but the DTC sats (magenta) are being operated at 360 km
Given that Starlink satellites can connect to some unmodified android phones, is there not a potential security concern there, that it could serve as a gigantic electronic signals snooping network with global reach, but American owned? I can't imagine other countries liking an overhead persistent Stingray system...
Breaking: Incoming changes to Direct 2 Cell Docket.FCC updates Status Date of SpaceX's application to 3/23/14 indicating an incoming update to the docket.
SpaceX seeks authority to operate its direct-to-cellular payloads in the lower altitude shells 340-360 km requested for its Gen2 system.
With tonight’s first Group 8 launch we kick off deployment of our commercial direct to cell constellation! @Starlink
SpaceX@SpaceXFalcon 9 launches 21 @Starlink satellites – including six with Direct to Cell capabilities – to orbit from California following the first successful demo of Direct to Cell texting in early 2024