Order Tray | Contact Us | Home | SIG Lists

[aprssig] Re: increasing WideX-X

Robert Bruninga bruninga at usna.edu
Fri Oct 1 23:47:06 UTC 2004


>> 1)  APRS is for local tactical real-time operations, not
>> statewide nets.
>
>I certainly believe that it started out that way,...- but my 
>observation is that the addition of Internet relaying, 
>IGATEs and findu.com have really changed that. 

Very true, they let people -voyer-to look over the shoulder
of local RF operations anywhere on the planet, and they do 
allow global APRS internet communications, but they don't 
change the laws of physics that say only about 60 or so 
people can reliably use APRS on RF (Ham Radio)...

>There appears to be a great interest in doing what I'll
>call long-distance tracking. My ... example is...trackers 
>in the backcountry south of Yosemite National Park,
>where digis are very sparse and IGATEs are ... typically
>four, [hops] away. 
> It's possible to get a tracker signal into findu from 
>Oakhurst, but the path needs to be at least WIDE3-3, [or] more.
>It's abundantly clear that this isn't what APRS was designed
>for. 

I disagree.  It is EXACTLY what APRS was designed for.
I cant think of a better application...   Let me explain, if I 
may to everyone listening...

Dont try to make an IGate solution to an RF problem.   You 
cannot violate the laws of physics.  It is very simple.

1) The APRS channel is 1200 baud. 
2) 1200 baud can handle about 60 packets per minute
3) But APRS depends on digipeaters, which double every
    packet. SO cut that to 30  unique packets per minute
4) Very-very few trackers are synced to Slot times, hence
    there is a 50% overlap average in your probability of a
    collision than your 1 second  packet, this cuts the 30
    to 20.
5) But there is more than one digi on the channel, lets say
    each digi can hear 2 others (in your sparse area), so
    divide the 20 by 2 and you get about 10 packets per
   minute under IDEAL conditions.
6) Remember that this is for ALL packets that EVERY digi
hears, and I bet those digis hear almost constant packet
QRM from out of area packets.  Over an hour, how many
stations do you see on your system in that area?  100?
at about 10 packets per hour per station thats about 16
packets per minute trying to get INTO your area that can 
only handle 10.

Your channel is theoretically totally jammed, but then
statistics say that occassionally you may get a packet
in edgewise.  Lets say your probability of doing that
was as GOOD as 30% (from a tracker in all that mess)

Now then you say you need to go 4 digis to hit the nearest
IGate.  The probability of that packet getting there is
30% * 30% * 30% * 30% or less than 1%!  At a tracker rate
of once per 2 minutes (to save power and be nice) this
means  your chance of being seen on an IGate is less
than once per 2 hours!

>But I think it's the reality of what [APRS] has become, or
>how people are trying to (ab)use it. In my part of the world,
>collisions aren't part of the immediate problem (though I'm
>sure they are at the end of the path!).

Unless you live on an island in the Pacific, or are on a
frequency different from 144.39, then I think you need
to reconsider.  "Collisions are 100%" of the reliability
problem, or the channel capacity limit on APRS on 144.39.  

Now, the correct way to do what you want to do is:

1) Put up enough digis to *hear* the RF from the trackers
2) Isolate the uplink frequency from the trackers from
    the statewide cross area traffic.  
3) Keep the number of trackers below  X for the intended
    area of coverage where X is based on how many trackers
    are trasmitting how often.  In the above example about
    Twenty 2 minute trackers could be handled on a given
    frequency.
4) Notice that this is even less than the usual ALOHA number
    of about 60 that you have seen me refer to, because the
    number 60, includes a mix of home stations, WX stations and
    only a small handful of mobiles and trackers...

And in conclusion,..... if anyone is still with me,.... that is why
it is time to consider trying to separate the local RF input of
digis from the statewide cross channel traffic on 144.39.
Check to see if 144.99 is available in yoru area, and start
adding alt-input digis so that you can have reliable RF reception
of APRS in your area...

thanks for letting me soap-box for a few moments on this
one...

de WB4APR, Bob





More information about the aprssig mailing list