Duplicates trams management ?

[Christian Barthod]

Hi team,

I'm trying to understand the APRS network and my question is about the duplicates frames management. 

If I assume on my picture, the WIDE parameter correctly adjusted on the 2 digi (WIDE1-1), how are managed the duplicates frames M2,M3,M4, transmitted by the mobile and received at the same time by other DIGI and iGates ?

73s Chris F8GHE

[Heikki Hannikainen]

On Sat, 22 Jun 2024, Christian Barthod wrote:

> Hi team,I'm trying to understand the APRS network and my question is about the duplicates frames

Duplicate packets having the same source callsign, destination callsign,
and packet data payload, are suppressed within a 30-second window.

This is done by the APRS-IS servers and digipeaters (and should not be
done by iGates, so that APRS-IS servers can see the duplicate packets,
and so that the duplicate packet dropping algorithm is implemented
consistently in a single place).

Because each APRS-IS server drops duplicates locally, and there are small
transit delays between the APRS-IS servers, two APRS-IS clients may get a
copies of the same packet via different paths (different digipeaters &
igates).

Sometimes packets may be delayed due to packet loss & TCP backoff, or
broken iGates (there have been problems with certain TNCs buffering and
delaying packets, for example). These may old packets to be received with
a delay, later than a more recent packet. aprs.fi has some additional code
to detect and drop these:
https://blog.aprs.fi/2008/03/on-duplicate-and-delayed-packets.html

There are also some recent issues with some LoRa iGates intentionally
corrupting packets by modifying the packet data payload. These cause
duplicate packets as well.

- Hessu

[Christian Barthod]

Thanks Hessu. So you confirm that the same frame received by several DIGI is correctly managed by aprsc (APRS-IS network) and that it would only be a few types of iGates which could be the cause of poor processing of duplicates?

73s Chris F8GHE

[Heikki Hannikainen]

On Sat, 22 Jun 2024, Christian Barthod wrote:

> Thanks Hessu. So you confirm that the same frame received by several DIGI is correctly managed by aprsc (APRS-IS network) and that it would
> only be a few types of iGates which could be the cause of poor processing of duplicates?73s Chris F8GHE

Yes, mostly it works alright. There are quite a lot of broken igates out
there, one way or another, but on the whole, it sort of works.

- Hessu