[aprssig] SATgates ? (again)

[John Wiseman]

Ok, Steve.

I'm convinced! Are you proposing this as a normal operating mode for an
IGate, or as a special SatGate mode? If the former, could the delay to
packets which are only seen directly be a problem, eg for messaging?

73,
John
 

-----Original Message-----
From: aprssig [mailto:aprssig-bounces at tapr.org] On Behalf Of Steve Dimse via
aprssig
Sent: 27 January 2016 14:02
To: TAPR APRS Mailing List
Subject: Re: [aprssig] SATgates ? (again)


> On Jan 27, 2016, at 6:06 AM, John Wiseman <john.wiseman at cantab.net> wrote:
> 
> Steve,
> 
> Although delaying directly heard packets for 10 seconds sounds a viable
> approach, I'm not sure you could guarantee that a packet will be
> retransmitted by the satellite and propagated to all Igates within 10
> seconds.

Actually, I think you can pretty much guarantee it except in the case of
malfunctions in the APRS IS. Understand it isn't that the digi'ed packet has
to get to other IGates, it has to get to the APRS IS, and specifically to
findU, aprs.fi, and other databases first. In the absence of a
malfunctioning station somewhere (see recent thread here) the latency in the
APRS IS is generally under a second.

Also, to be clear, the IGate software need not do any comparing for this to
work. It simply holds any directly heard packet (no used vias in the path)
for 10 seconds. Even if it hears the digi'ed packet itself, the digi'ed
packet will have a used digi in the path and will not be held. It gets send
immediately from the IGate to the hub it is connected to. When it releases
the direct packet that it is holding that first hub will filter it. If
another IGate hears the digi'ed packet it sends it with no delay, and so
there is 10 seconds for it to travel the APRS IS and reach all the databases
before the first IGate releases the direct packet. 

The APRS IS functions with remarkable reliability. And if it is functioning
correctly the 10 second delay will work. Are there situations that could
still result in a direct packet getting into findU before the digipeated
one? Yes, but very rarely and only in the case of malfunctions.

This scheme will allow all the direct packets to be heard, the extra 10
second delay is no problem, and will let the digipeated packets have
preference the vast majority of the time. This is a far better solution than
a filter of direct packets with a list of blessed stations. It is easy to
implement and does not need maintenance as new satellites come online.

Steve K4HG
_______________________________________________
aprssig mailing list
aprssig at tapr.org
http://www.tapr.org/mailman/listinfo/aprssig