Erlang?

[Richard Barth]

Alex,

If I don't misunderstand your question, you're asking about Erlangs
generally rather than specifically about their use in radio channels.  

As pointed out on the web site whose URL you gave, erlangs were
first created for use in the telephone biz.  The idea was that if there
are 100 people in town A and 100 in town B, you don't need 100 lines in
the trunk between the two because chances are they won't all be talking 
from one town to the other at the same time.  How many do you need?  You
can look count the amount of actual traffic during the busiest hour and
decide how much risk you will accept that during some future busy hour
callers will get busy signals, and also decide how long you want a user
to wait before his call goes through.  Then you use the models.

The theory is useful anyplace you have a bunch of people waiting to use
a limited resource: telephone trunks, microwave channels, paging systems
and, as pointed out earlier, bank clerks.

Another way of measuring traffic density is in hundreds of call seconds,
or ccs.  You'll find this in the literature as well.

73,

Dick

Bob Bruhns wrote:
> 
> Hi Alex,
> 
> I saw Erlang analysis used in the paging industry, to quantify the
> probability of congestion.  In paging, congestion is deadly because
> if the system is clogged and pages are delayed in a que, users will
> start resending en-masse, and the system will REALLY choke.  I saw a
> 12% increase in system capacity result in the elimination of 20
> minute (!) paging delays on Friday evenings.
> 
> It was very confusing to do any kind of system test in those
> conditions, if you did not realize what was happening.  One Friday
> evening I was trying to get the TDD system working.  I was trying
> different configurations, but test pages did not come through, so I
> thought the configurations were wrong.  Then about a half hour
> later, the first of the test pages started coming through!  Geez.
> 
>   Bob
> 
> ----- Original Message -----
> From: "Alex Fraser" <beatnic@comcast.net>
> To: <tacos@amrad.org>
> Sent: Friday, May 17, 2002 11:26 AM
> Subject: Erlang?
> 
> > Are Erlangs ever used in radio channel analysis? It is new to me.
> >
> >  http://www.erlang.com/whatis.html
>