[aprssig] HF APRS new ideas

[Lynn W. Deffenbaugh (Mr)]

Bob Bruninga wrote:
> What we need are SMART HF 2-way IGates that will ONLY consider theseleves to be the "local" Igate for relaying that message if and only if IT heard the mobile in say ...the last say 10 minutes.  This would go a long way towards eliminating dupes from multiple two-way IGates.
>
> Only the one with immediate NOW propogation to the mobile would send the messge.
>
> Can someone comment on how well something like this can be managed with most existing IGate packages?  Can the statistics of their "HEARD LOCAL" list be used smartly?
>   

I'm designing ARPSISCE/32 to do exactly that, including much more 
granular configuration of digipeating, both within and across individual 
RF interfaces.  Configuration of the definitions of "recent" (time) and 
"local" (hops) for gating messages to RF are in the plans, probably 
sooner than later.

However, the propagation of HF, at least the 30m that I've been 
monitoring, is interesting and unpredictable to say the least.  I'm 
fairly certain that coordination of which Igates are even configured to 
gate messages from -IS to RF will be required to avoid just the 
collisions you describe.  And even then, the "hidden transmitter" 
becomes a major issue if a station in the center of the US transmits and 
is heard by an east and west coast IGate and then a messages appears on 
the -IS for that station.  Both of the gates will consider the station 
local and will transmit, probably completely eliminating the ability for 
the station in the center to receive it.

I've been trying to think of a zoned approach for IGates, possibly 
encouraging an east-to-west priority or vice-versa, to at least stagger 
the gating of the packets since on HF it's unlikely that the gates can 
hear each other and the FM capture effect isn't going to help the receiver.

Lynn (D) - KJ4ERJ - Author of APRSISCE for Windows Mobile and Win32
Monitoring 30m APRS on both AX.25 and PSK63 and Receive-Only IGating as 
KJ4ERJ-2