GregS wrote:

In article <T1128978093@djwhome.demon.co.uk>, david@djwhome.demon.co.uk (David Woolley) wrote:

In article <434A3B3D.9080606@nowhere.com>, "Dave <nospam@nowhere.com>" <>
wrote:

2) Much higher S/N, to decode, so even higher powers would be needed. I don't know, but I doubt you could do it with a S/N or less than 30 dB.

What you also have yet to account for is that the signal is reflected from
many different ranges so is convoluted in time by the distribution of
reflector by range.  Any modulations scheme with a baud rate (in the
strict sense) greater than the order of about 150 baud is going to
suffer severe inter-symbol interference (I've used 1000km for the
moon's radius, rather than looking up the correct figure).

This makes sense as I was reading it. But, I just remembered we used
to range the moon using digital correlation. I can't remember the frequency
of the moulation or type of modulation, probably PM. There were also tone
ranging devices which were set up so the moon could also be ranged.
They used multiple tones for getting locked.
The NASA 2300 Mhz frequency was used. I worked at Appolo site
at Goldstone.

greg

An interesting point. Digital TV (8VSB) signals are ideal for
moonbounce IF you don't want to decode, just receive. Since they
are mostly random sequences (with some pseudorandom parts) with
a flat frequency spectrum they are ideal for doing autocorrelation
analysis. The minimum period is about 80 nsec and the maximum is infinite. The powers are very high. OF course, you would want
transmitter and receiver to be far enough apart that the
moon was near the horizon at the transmitter and near the zenith at
the receiver.

Doug McDonald