[aprssig] Re: Fixing Los Angeles

[Robert Bruninga]

>>> "Jim Lux" <jimlux at earthlink.net> 01/08/05 3:28 PM >>>
>There are some good ideas there, [on Fixing LA]...
>It would be almost impossible to put up a single digi...

Thanks for taking a look at it...  Referring to a single digi
was only hypotehtical to start the concept thinking process...

>Likewise, I'm not sure that "This will work, because the downlink 
>from a DIRECT digi is orders of magnitude more successful 
>in getting to a mobile on the ground with little QRM" is actually true.
>It's not usually a matter of 3 or 4 dB or even 10 dB, it's a matter 
>of path attenuations of 20,30, or 40 dB greater than the
>line of sight distance would imply.  It's not just QRM, it's
propagation
>dead bands.

THanks, jim.  That is not what my  sentence was about.
That sentence was saying that considering the path between
a mobile and a digi, the mobile (who only hears 5% of the
QRM that the digi hears) will hear the digi much more reliably
than the digi will hear the mobile (over *any* given path).
It was not meant to be a comment about geography at all.

What it meant was, dont use how often you are digipeated
as a measure of an RF path because how well the digi hears
the mobile is confounded totoally by the QRM issue.  Instead,
evaluate that RF path for planning purposes based on the
mobile hearing the digi which is the better measure while
minimizing the efffect of the totally saturated QRM at the
digi end of the link.

>It's very, very difficult to find a "small" number of sites that 
>cover ALL of LA.   While Bob's page talks about using 3 high 
>digis, I seriously doubt whether 3 could actually get the 
>coverage required.  His map shows N6EX-2, W6SCE-10, 
>and WB6JAR-10, but in reality, I doubt that those would 
>actually cover even 50% of the territory shown on the map.

I should have removed the callsigns, since I was using digis
from an old file more than 5  years old.  You guys know best
what very high digis would be best to consider.  I also did
not mean to imply that 3? was a magic number.  I used the
question mark in all cases so that you guys ccould tell
us what the right number is...

>In a distance of substantially less than a few miles, on major 
>roads, you may move in and out of the intersecting coverage 
>area of several digis multiple times, so if all you're limited to
> is what's in view of the digi you can currently see, you're 
>going to miss a lot of stuff that's actually quite close.

Yes, but since the overlaping competing coverages you describe
above are on the same 144.39 fequency, then again, which digi
is 10 dB stronger at any instant *does* vary very much and 
causes multiple flips back and forth between digis.  And all it
takes is that 10 dB change to flip between digis becaue of
the FM capture effect.   But in the ultimate cross-band to UHF
scenario on different freqs, then the signal may get 10 dB 
weaker, but there is no interfering digi on the sam fequncy
that will capture your discriminator.  Thus you may have 20, 30
or even 40 dB of additional margin before you lose a packet.

>Now, if you could somehow get the dozen or so digi's that 
>would span most of LA to cooperate, and run all incoming 
>packets to a central clearing house before retransmitting 
>them...

Nah, as you said, too complex, not KISS and you would never
find the cooperation.  It would never happen...

>However, this is a fairly challenging(!) system design and 
>implementation problem.  Who's going to coordinate this?
> Who's going to run the "master" router?  Who's going to 
>be standby or backup? What happens when someone
>takes a digi down for maintenance, or gets tired of running it, 
>or can't get <snip>

Based on that paragraph I think you completely missed the
entire concept.  There is zero coordination required.  There
is no master site, there are no crosslinks, no new hardware
nothing.  Just each digi owner deciding if he want to do it
at his site or not.  If he decides, he simply does it and it
can all be done by REMOSTE-SYSOP by simpl;y changing
only 3 parameters.  The UIDIGI list, the UIFLOOD parameter
and the UITRACE parameter.  Done.   No existing user
loses anything, everyone in his foot print gains.  One digi
at a time...

Bob