Order Tray | Contact Us | Home | SIG Lists

[ax25-layer2] ECC vs. FCS

Pete Loveall AE5PL pete at ae5pl.net
Fri Aug 4 13:28:45 UTC 2006

> -----Original Message-----
> From: Wes Johnston, AI4PX
> Sent: Thursday, August 03, 2006 10:42 PM
> To: Discussion of Link Layer use of AX.25
> Subject: Re: [ax25-layer2] ECC vs. FCS
> I realize that this discussion concerns other uses of ax.25, 
> but with regard to aprs.... something that was suggested for 
> a sort of pseudo FEC was a doppleganger packet.
> A normal TNC would have an opportunity to hear each copy of a packet.
> If it heard the first copy it would digipeat that copy once 
> the channel was clear.  If it heard the 2nd copy, the dupe 
> filter would prevent _that_ copy from being digipeated.  If 
> it missed the 1st copy there's a chance the 2nd copy would be decoded.

The problem is loss of synchronization.  TXDelay is there to allow the
receiver to synchronize with the transmission.  While mobile picket fencing
might be overcome, the more common "hidden" transmitter effect wouldn't be.
Something to keep in mind.

This is getting a bit off topic as it is focusing on layer 1 and application
implementations.  While interesting, it has nothing to do with the AX.25
protocol since there shouldn't be included in the specification requirements
for individual implementations.  Requiring duplicate packets would be an
implementation, not a standards issue.

Implementing extended frame lengths, modulo 128, etc. are AX.25 protocol
issues and are being discussed here (and have been discussed extensively).
My post is not an admonishment, simply a "let's keep on track" encouragement
as I am seeing this thread starting to digress into a non-layer 2


Pete Loveall AE5PL 
-------------- next part --------------
A non-text attachment was scrubbed...
Name: smime.p7s
Type: application/x-pkcs7-signature
Size: 3097 bytes
Desc: not available
Url : http://www.tapr.org/pipermail/ax25-layer2/attachments/20060804/e5060d39/attachment.bin 

More information about the ax25-layer2 mailing list