Anthony,

If your interested in buying out a solution here is a radio transceiver
ready to interface to a uC and already FCC tested.

http://www.ozemail.com.au/~martech/indigo.htm

I haven't used one yet, but I looked into it for a project that's on the
back burner.  This unit would allow two way communication which would allow
you to eliminate contention via an ACK or polling from the master unit.
After reading this post, I think your definitely going to have to do
something about contention.

Here are a few specs from the website :

Freq : 916.5 MHz
Op Voltage : 4.5 - 5.5
Op Current : < 18 mA ( during TX or RX )

The only possible gotchas are that I'm not sure about the range and the
operating temp is -40 - 85C.  If I remember correctly, unit prices are
around $90 in quantities < 10. Price drops to roughly $60 @ 2500+.

On Monday, July 12, 1999 8:59 PM, Anthony Clay [SMTP:adctemp@hotmail.com]
wrote:
> Thank you Greg and Adam, your descriptions helped greatly.  Here is a
better
> description:
>
> The system is to be designed to present a computer to figure out which
> heffer cows are in heat.  The transmitters are located on the backs of
the
> heffers, and when the "dummy steers" try to mount them, it is a prime
> indication that a cow is about to come into heat.  The computer comes in
> here, it can take the cows history and come up with all sorts of numbers.
> The microprocessor is used (instead of a simple beacon) to provide a
single
> component that can debounce the switch, keep an internal history as a
backup
> for that cow, and most importantly, protect against contention.
>
> Contention does not seem to be likely, but it is a curious issue.  If
this
> system is properly used, only the females up for heat, as calculated by
the
> computer, would be kept in the "holding area",  the "holding area" would
be
> closely monitored by the computer for "mounts".  Females in the holding
> area, however, are more likely to be mounted at or near the same time,
very
> large ranches can have several hundred cows in holding.  The ranchers
can't
> watch them all, but the computer can.  Mounting usually occurs around
peak
> times, about 1 to 2 hours of the day, further increasing the chance of
> contention.
>
> Lost data packets, however unlikely,  could easily cause the computer to
> miscalculate, thereby corrupting, albeit slighty, the history of the cow.
> The bottom line is missed "heat" calculations, a possible $10,000 cow per
> missed packet does not cost money, but I would like to know about it.
>
> Legal Question:  "Perodic Operation," as stated by the FCC in Title 47,
> Part 15 Section 23; does allow the transmission of an identifing signal.
> Can a data packet still be included?
>
> Anthony