3 ms·
> you could get time signals from GPS (or GLONASS) You forgot to mention Galileo ... GLONASS may not be politically attractive :) But beyond the GNSS ecosyste
by traceroute66 3y ago
> you could get time signals from GPS (or GLONASS)
You forgot to mention Galileo ... GLONASS may not be politically attractive :)
But beyond the GNSS ecosystem, there are of course other interesting options:
Safran STL[1] which uses LEO sats. Apparently it works well in places where you can't get a good GNSS signal (i.e. indoors without an external antenna). (This was previously Orolia STL, but they were acquired by Safran).
Most national time labs also offer a leased-line service, e.g. NPL (UK)[2]
There is also a very niche (read: VERY expensive) commercial timing-as-a-service product from a company called Hoptroff[3]. You license both the service/connection plus their software clients, so the $$$$$$ add up. Definitely one of those "if you have to ask you can't afford it" vendors.
[1] https://safran-navigation-timing.com/product/stl-option https://safran-navigation-timing.com/product/stl-option
[2] https://www.npl.co.uk/npltime https://www.npl.co.uk/npltime
[3] https://www.hoptroff.com/ https://www.hoptroff.com/
- jjtheblunt 3y agoGalileo too! (Pre coffee oversight) And, recursively, from celltower signals for any CDMA-related technology, i think.