[aprssig] timeslotting on HF ? was > 15. APRS trackers on10m (Robert Bruninga)

[Dave Baxter]

Certainly keeps to the KISS principle...

The last 4 minutes could be carved up between ongoing emergency traffic,
weather, system announcements, and ????   "Ongoing" emergency traffic,
being anything but the first emergency "Call"...

You could distribute the 4 "spare" minutes arround the hour, every 15
for instance, would probably be better for weather and emergency stuff?

As you say, and perhaps with a little "guard band" of perhaps say +-2.5S
at each minute boundary (5 second total) it should be easy to syncronize
manualy, or with very simple (embedded) software, perhaps even by
monitoring ongoing traffic, and reverse coding from the callsign's
heard..  Getting not so simple now, but.....

If you make the guard band larger (10 sec total?) that's plenty of time
for a genuine emergency shout, to be heard mostly in the clear, and able
to repeat every minute, until an ACK of some sort is returned, perhaps
by an I-Gate?

What's needed is an idiot proof way of synchronising to a universal time
source, GPS, Internet, WWV, Rugby etc or whatever?

Dave G0WBX.




> -----Original Message-----
> From: aprssig-bounces at lists.tapr.org On Behalf Of David Rush
> Sent: Monday, November 28, 2005 3:02 PM
> Subject: Re: [aprssig] timeslotting on HF ? was > 15.
  APRS trackers on10m (Robert Bruninga)
 
> How about a VERY simple time-slot algorithm... such as the 
> last letter of your callsign indicates the minute past the 
> hour or half hour within which you're allowed to transmit?  
> Calls ending with A could transmit during
> hh:00:00-hh:00:59 and hh:30:00-hh:30:59, calls ending with B 
> could transmit in hh:01:00-hh:01:59 and hh:31:00-hh:31:59, 
> and so on.  The last four minutes of each hour or half hour 
> could be reserved for something special, or simply a free-for-all.
> 
> Each user would have two one-minute slots per hour to work 
> with.  Such a long slot accommodates manual 
> timeslot-operation if necessary, and makes it easier to keep 
> track of time, especially for small, PIC-based trackers that 
> have very limited code space.
> 
> Within one's appropriate minute, users could pick a random 
> time to transmit (or sub-timeslot within a local area if 
> multiple regular share the same timeslot).
> 
> As far as I know, this should work equally well for most 
> international callsign patterns, but I could be wrong.  
> Callsigns ending with a digit would use the last-most letter 
> in the callsign, maybe?
> 
> I first thought of an even simpler algorithm, where the 
> (first) digit in the callsign aligns with the last digit of 
> the current minute, but then realized that most callsigns in 
> a given area would end up with the same timeslot (at least 
> the way calls here in the US are assigned).
> 
> Just thinking out loud, and knowing that simple is often a 
> good thing...
> 
> David, KY7DR