Order Tray | Contact Us | Home | SIG Lists

[aprssig] APRS low-power-local ALT input channel

Robert Bruninga bruninga at usna.edu
Sat Sep 25 03:50:22 UTC 2004


>>> HamLists at ametx.com 09/24/04 2:12 PM >>>
>Per previous discussions, you are promoting transmitting 
>on a shared frequency without ANY form of collision avoidance. 
>I see [this as] ... the encouragement of bad operating practice 
>and bad network design.

Ah, that is a common missunderstanding.
When any system uses HIGH site repeaters to serve many ground
level users, the resulting network is not really a carrier sense
multiple
access system.  It is almost pure Aloha beacuse almost none of the
users can hear each other, and so when they transmit, they have
no way of avoiding a potential collision simply by listening.

Anyone who doesnt belive this should simply look at his APRS
screen and count the number of other usres he hears direct.
It is probaby significanlty less than 1%.  
 
For example, at my average house on average terrain with an
antenna at 80 feet above ground half way between Baltimore
and Washington DC (some of the highest density APRS around), 
I just checked and of the 306 stations I have heard in the last
4 hours, I only hear 9 direct, and 4 of those are digipeaters.
Of the 5 other users, only 1 is a home, 2 are WX stations and
2 are mobiles that at some time in the last 4 hours passed within
a few miles of my location.

Work the numbers and how often those other users transmit
and you see that my chances of colliding with one of those
5 stations that I can hear is 1 chance in about 300.  Or less
than a third of 1%.  So no matter how hard I listen, 99.3 %
of the time it does no good.

Now for the ALT channel that I recommended so that low power
trackers can be heard, remember that these trackers only transmit
less than 1 Watt.  Hence even for that 0.003 of the time when they
might actually hear something, who is going to win?  The fraction
of a watt low power tracker or the 50W home station or mobile?

But who else is on the low power alternate channel?  Only other
low power trackers.  Lets say that maybe 5 of them are in range
of this alt-channel input.  And they are beaconing at once every
2 minutes.  Their chance of colliding is 4%.  Their chance of
being heard is 96%

Now what do they get for this CHOICE they have made to
oeprate on this alternate low power input channel compared to
transmitting on 144.39?:

1)  On the ALT channel they get a 96% chance of success and 
    only a 4% chance of collision on the ALT channel
2) On 144.39, in a busy area, their less than 1 watt signal has 
    about a zero chance of success and about a 100%
    chance of a collision with 305 other stations and digis.

It doesnt take much to figure out which of these selections
is the proper network design choice.

>There is nothing inflammatory about this statement. 
>It is bad operating practice to transmit on a frequency without 
>listening first to avoid causing intentional interference. It is a bad 
>data network design to have data devices share a medium 
>(in this case the alternate frequency) with no collision avoidance 
>mechanism.

That statement is quite illogical in view of the facts.

>A far better solution, IMO, is to educate people about excessive 
>paths and beacon rates, and to implement restrictions at the 
>digipeaters such as not digipeating packets that have excessive paths.

Again, another common missunderstanding.
That will do nothing to change the ALOHA network to a CSMA
network.  No matter what you do with  paths, it is not going
to change the fact that any given mobile cannot hear the other
99.3% of the other users that he is sharing the channel with.

Pete Loveall AE5PL
mailto:pete at ae5pl.net 

> -----Original Message-----
> From: Robert Bruninga
> Posted At: Friday, September 24, 2004 11:20 AM
> Subject: [aprssig] APRS low-power-local ALT input channel
> 
> Establishing a local ALT input for APRS in USA:

_______________________________________________
aprssig mailing list
aprssig at lists.tapr.org
https://lists.tapr.org/cgi-bin/mailman/listinfo/aprssig





More information about the aprssig mailing list