Topics

UNID in W Scotland

david
 

Hello knowledgeable folks,

Subject line changed for clarity.
My apologies for the delay. We have had visitors, and the 'lady of the house' imposed other priorities ;(

I think this is the record for the UNID that caused John so much frowning. I hope it gives up its secrets to those who can decode it:

TIME: 2018-08-29 04:28:05 FREQ: 2187.5 DIST: -- Km
SYMB: 120 120 056 030 004 020 000 108 000 026 030 078 000 118 126 068 ~~~ ~~~ 126 ~~~ 126 122 094 122 122
FMT: SEL
CAT: SAF
TO: SHIP,563004200,???
FROM: COAST,002630780, UNID
TC1: TEST
TC2: NOINF
FREQ: --
POS: --
EOS: ACK
cECC: 94 OK

RAW FEC: 125 108 125 107 125 106 120 105 120 105 120 104 056 120 030 120 004 056 020 030 000 004 108 020 000 000 026 108 ~~~ ~~~ 078 019 ~~~ 030 118 010 126 000 ~~~ ~~~ ~~~ ~~~ ~~~ 068 126 ~~~ ~~~ ~~~ ~~~ 126 122 ~~~ 094 126 122 122 122 094 ~~~ ~~~ ~~~ 043


Best Wishes
David, GM8XBZ

-----Original Message-----
From: dsc-list@groups.io <dsc-list@groups.io> On Behalf Of John Pumford-Green
Sent: 04 September 2018 07:54
To: dsc-list@groups.io
Subject: Re: [dsc-list] REU _ DSC Logs from W Scotland


David's UNID 002630780 confused me when I noticed it a couple of days
ago, and still confuses me now!

It looks as though the ECC (Checksum) in David's message calculated
correctly even though there's obviously an error in one of the MMSI
digits (78 instead of 10).

This just goes to show that the error detection mechanism of DSC isn't
foolproof. It would be interesting if David could find his copy of the
message in YaDD and copy/paste the symbols so we can compare it to what
Peter and I received. I'd like to see if I can explain why the Checksum
in David's copy was OK despite the error.

Cheers,

John
GM4SLV

John Pumford-Green
 

Thanks David,

That confirms my suspicion that your message was riddled with errors even though it appeared error free!

The "real message" as received by Peter and myself:

120 056 030 004 020 000 108 000 026 030 010 000 118 126 126 126 126 126 126 126 122 094

Your copy:

120 056 030 004 020 000 108 000 026 030 078 000 118 126 068 ~~~ ~~~ 126 ~~~ 126 122 094


The received ECC value from both versions is the same : 94

Your message, even with the errors, gives a calculated ECC value of 94, and thus matches the one received.

The errors:

the 4th MMSI symbol from the sender : 078 (should have been 010)

the 1st symbol from the "additional data" section : 068 (should have been 126)

The 2nd, 3rd and 5th symbols from the "additional data" section : ~~~ (should have all been 126)

"~~~" indicates that the symbol failed the initial parity check and has been discarded. When calculating the ECC value these "~~~" symbols are treated as "zero" by YaDD (which I think is the wrong thing to do...)

By a strange quirk of fate the errors, 78 instead of 10, 68 instead of 126, and "0" instead of 126, means the ECC still comes out to 94.

And because the parity fail errors are in parts of the message that carry no information they weren't visible in the "decoded" version of the message either, so there was no clue.

This goes to prove that DSC is still prone to false positives. I've long-since held the view that YaDD should discard messages if there are ANY parity errors (symbols with "~~~") regardless of whether the ECC check is successful....


Cheers,

John

On 06/09/18 11:08, david via Groups.Io wrote:
Hello knowledgeable folks,
Subject line changed for clarity.
My apologies for the delay. We have had visitors, and the 'lady of the house' imposed other priorities ;(
I think this is the record for the UNID that caused John so much frowning. I hope it gives up its secrets to those who can decode it:
TIME: 2018-08-29 04:28:05 FREQ: 2187.5 DIST: -- Km
SYMB: 120 120 056 030 004 020 000 108 000 026 030 078 000 118 126 068 ~~~ ~~~ 126 ~~~ 126 122 094 122 122
FMT: SEL
CAT: SAF
TO: SHIP,563004200,???
FROM: COAST,002630780, UNID
TC1: TEST
TC2: NOINF
FREQ: --
POS: --
EOS: ACK
cECC: 94 OK
RAW FEC: 125 108 125 107 125 106 120 105 120 105 120 104 056 120 030 120 004 056 020 030 000 004 108 020 000 000 026 108 ~~~ ~~~ 078 019 ~~~ 030 118 010 126 000 ~~~ ~~~ ~~~ ~~~ ~~~ 068 126 ~~~ ~~~ ~~~ ~~~ 126 122 ~~~ 094 126 122 122 122 094 ~~~ ~~~ ~~~ 043
Best Wishes
David, GM8XBZ
-----Original Message-----
From: dsc-list@groups.io <dsc-list@groups.io> On Behalf Of John Pumford-Green
Sent: 04 September 2018 07:54
To: dsc-list@groups.io
Subject: Re: [dsc-list] REU _ DSC Logs from W Scotland
David's UNID 002630780 confused me when I noticed it a couple of days
ago, and still confuses me now!
It looks as though the ECC (Checksum) in David's message calculated
correctly even though there's obviously an error in one of the MMSI
digits (78 instead of 10).
This just goes to show that the error detection mechanism of DSC isn't
foolproof. It would be interesting if David could find his copy of the
message in YaDD and copy/paste the symbols so we can compare it to what
Peter and I received. I'd like to see if I can explain why the Checksum
in David's copy was OK despite the error.
Cheers,
John
GM4SLV