Real time active aiming?( was Re: [Laser] Laser comm stuff)

[Derek Weston]

> One reason is if you loose lock either way, you have to start from scratch.

True. 

> You maintain lock by a very small dither in pointing, and re-acquire by a
> raster or conical scan, looking for a maximum return signal.

The same idea can be used with the alignment method I've described,
periodically interspersing data packets with alignment test packets at
slight offsets. The end receiving the test packets can return the "best"
coords for the sender on the basis either of signal strength (but that
may need more complicated electronic hardware) or the average of the
coords at which test packets were received.
There'd be some overhead involved but I expect it would be insignificant
as alignment drift is generally slow.

> Each end re-acquires stand alone.

True again, and I agree that for some reason that seems more appealing.
However you pay a price in HW and the real benefit in average data
transfer rate would I expect be small.

> Clearly, beam size is an issue. For max. received power, the beam should under > fill the receive aperture. For retros, it needs to be a bit larger.

Sure, but the tighter the beam the faster the possible data rate, all
else being equal.