60 live Starlink satellites over Aberdeen: strongest-signal and cheapest-to-sync agree ~88% of the time, because strong signals come from slow-moving satellites.
Hi all,
This week started with what felt like a clever idea.
Over the last few weeks I've been measuring what it costs a 5G satellite terminal to stay in sync with a LEO satellite whipping across the sky — it has to keep refreshing a timing and Doppler correction, and a low satellite (drifting fast) costs a lot more to keep up with than a high one (drifting slowly).
So here was the thought: when the phone can see several satellites at once, it normally just grabs the strongest signal. What if it instead grabbed the one that's cheapest to keep in sync? Could save battery and signalling, right?
I built it, pulled 60 live Starlink satellites over Aberdeen, and let three different handover strategies fight it out.
The clever idea mostly evaporated — and in a satisfying way. It turns out the strongest satellite and the cheapest-to-sync satellite are usually the same satellite.
The reason is nice and physical: a satellite gives you a strong signal when it's high overhead and close — and that's exactly when it's drifting slowly across the sky, which is what makes it cheap to track. Strong and cheap come from the same geometry. About 88% of the time, the two strategies made the identical choice.
In the small fraction where they disagreed, the "cost-aware" version did shave off about 9% of the sync refreshes — but it paid for it with a weaker signal. And the obvious tweak — just stay on a satellite longer to avoid switching — actually made things worse, because you end up clinging to a satellite as it sets and gets expensive.
So the honest result: the boring old "strongest signal wins" rule is hard to beat. Which is genuinely useful to know — it tells you not to over-engineer this part of a satellite phone.
This one quietly pulled together a lot of the earlier weeks — the link budget from Week 1, the handover scheduling from Week 3, and all the pre-compensation work from Weeks 5 to 7 — into a single decision. Eight down, four to go, and the flagship's coming into view.
Full write-up on Medium: https://medium.com/@shaifiqbal/should-a-satellite-phone-chase-the-strongest-signal-or-the-cheapest-one-to-keep-in-sync-279450ffb25e
Code, tests and plots on GitHub: github.com/shaifiqbal/ntn-precomp-aware-handover
More next week.
Shaif