Re:Modeling Starlink Capacity
by cycomanic on Sunday October 09, 2022 @01:02AM
(#33138408)
Something not covered in the article, but I have wondered about. How does Starlink do the handover? The satellites are in Leo and thus have orbit times of around 90min or so (not sure about the exact number), so every cell must be served by several satellites at the same time no? Otherwise we would see significant interruptions at the handover.
[ Parent
| on HN ]
I assume the satellites are synced so the handoff could be scheduled in advance and done in milliseconds. All you'd notice is some jitter.
I don't think it's any harder than driving on a highway where you'd do a cellular handoff every two minutes or so.
well, some (or all?) have laser links now, so as another sat gets covisibility with a ground station and user terminal, it can just instantly pass the traffic off
without laser link, it's looks to me that most of the time in populated latitudes, there will be more than one sat's beams covering a cell, and if not then there would be interruption in coverage
the terminal should handle transmitting to multiple sats that are close by, but i dunno how omnidirectional the dishy is
Re:Modeling Starlink Capacity
by viczsaszbo on Sunday October 09, 2022 @01:55AM
(#33138578)
The beam hopping (TDM) part of the article is similar to satellite handover. There's a lag before a beam returns to a location or a new beam is over a location, and the beams can't overlap (nco=1).