Order Tray | Contact Us | Home | SIG Lists

[aprssig] ALT input channel - different perspective

Wes Johnston wes at johnston.net
Sun Sep 26 13:20:34 UTC 2004


I know this alt input thread is going to turn into a holly war soon..... ;-)

It seems that there are two camps... those who are worried about 
widespread deployment of these alt input digipeaters because their 
concept involves transmitting in the blind on 144.99.  The other camp 
says it's OK to TX in the blind on that frequency as long as it's 
locally coordinated. 

I want to throw another idea in.... that of the mixing of CSMA and Aloha 
type collision avoidance. 

In order for CSMA to work all stations in a network have to be able to 
hear all other stations.  This works very well, but not perfectly, on a 
wired network.  I still see collisions on my home ethernet, but not as 
many as we have in packet.  The number of collisions you will see in a 
CSMA network becomes a function of the overlap in coverage area that the 
network participants have.  The greater the overlap, the fewer collisions. 

Let's consider two neighboring digipeaters with an effective range of 30 
miles.  If they are 30 miles apart they share an overlap of just a 
little less than 50%.... so there is a little more than 50% chance one 
could be hearing something the other can't, and the other might 
transmit.  That's OK - well fair --- livable.

Now consider a car with an effective range of 5 miles.  The car can hear 
stations in a 78 square mile area around him.   The digipeater can hear 
stations in a 2800 square mile area.  The car doesn't hear 97.5% of the 
digpeater's coverage area.  That means that there is an _enormous_ 
chance that the car will try to TX when the digi is hearing another 
station that the car can't.  In this case, the car getting into the digi 
is a crap shoot... err I mean aloha. 

My point here is that we are presently operating two network models at 
the same time on 144.39.  Some of the participants in the network are 
running CSMA while others are running aloha.  So let's coordinate some 
local frequency (ie 144.99) and let the aloha stations go there.  From 
there, their packets will enter the CSMA network model properly.   Blind 
stations like tiny trak and pocket trak AND mobiles who can't hear a 
significant portion of the digipeater's coverage area should consider 
themselves "aloha" stations and take their chances on 144.99 - where 
they compete for access with a very limited number of stations.

Wes




More information about the aprssig mailing list