[time-nuts] time nuttery in Space Communications

[jimlux]

One thing in all these DSN histories is that they don't make very much 
of the essential thing that separates Deep Space radio links from Near 
Earth radio links, and that's the timing: specifically the coherent link 
between the received signal and transmitted signal.  I suspect that's 
just because it's an "of course it's coherent" for everyone who does 
deep space comm, so it's of no great notice.


One thing I find in folks building and using ground stations (and 
spacecraft) for LEO (vs beyond GEO) is that they essentially treat the 
communications path to the spacecraft (telecommand, in the lingo) as 
entirely separate from the communications path from the spacecraft 
(telemetry, in the lingo).

This is also found in textbooks on communications systems at whatever 
level - you have a sender, you have a propagation path, you have a receiver.
If there's a "connection" between the two directions, it's usually 
handled as a "network" thing (in the sense of ARQ or ACK/NAK, perhaps).

Things like Doppler and oscillator stability are "nuisances to be 
compensated for" so that the receiver can be tuned to the right frequency.

More than one ground station system for Earth orbit has no idea how long 
it takes for the signals to propagate through the chain - it doesn't 
matter - I put bits in at one end, and not too long (milliseconds, maybe 
seconds) the bits come out the other end.


However, for those of us using the radio signals to track and navigate, 
or to do science, that "round trip light time" or "phase difference 
between receiving the signal at two different stations" is really, 
really important.

And, a bit of codger, "get off my grass" sense (and I'm a relative 
newbie, having only done this for 20 years) - all those folks who are 
excited about going to the Moon are having to learn this all again. 
There's no GPS to give you your position, an omni antenna won't close 
the link, etc.