[SCCC] Fw: Anti-Skimmer Petition

Tue Apr 29 18:28:17 EDT 2008

Hi Art,

OK, I jumped in not because I'm advocating using it for contests, but rather the petition has silly paragraphs which make it impossible for anyone to take seriously and so I decided to point those out.  If you are going to send out a petition, don't make it simple for the other side to poke it full of holes...

To your point on error correction - my bad - I'm mixing error correction and noise resistance.   In the case of PSK the encoding is via phase reversals, it's not error checking per se, but aids in the DETECTION of low level signals and renders it much more robust in the face of noise - most noise doesn't involve phase reversals, whereas in CW the noise looks just like the signal itself (spikes of signal).  I agree that CW is also digital, but not all digital modes were designed for the same thing...

I've never used CW Skimmer but have used the reader in Mixw for CW (pretty awful), but the same reader happily decoded PSK-31 from South Korea a few weeks ago - I called CQ on 17M, heard nothing but then looked and saw my call coming back from DS5RYB - I almost missed him since it was completely silent and I didn't even see a waterfall trace!

Just think about what is needed for CW:

Decoding CW REQUIRES picking up the PEAK TOPS and the LENGTH of those peak tops (to interpret dits and dahs) - impossible if the signal is in the noise which looks like a LOT of dits with random spacing.   It would be impossible to fish out a signal at the noise floor but it can be done in PSK-31 due to the difference in encoding (phase reversals).   Weak signals are slightly above the noise floor but noise has periodic peaks that jump up higher than the baseline - the closer to the noise floor, the more often this happens.  Thus, one gets an extra dit inserted into a letter, how does the computer figure out which one was the extra or if one was extra at all?   Also, you have to pick a peak level to sample the peak length - a sharp but tall peak will look like a dah near the base and a dit at the top.   Sampling at the top of that peak may not work since the signal may drop and the other peaks will be below the sampling threshold...

CW is beautifully simple, but that same simplicity confounds decoding near the noise floor, while you do an awful lot of complex reasoning to pick out those weak calls in your brain.  Even you can not pick out a signal that is AT the noise floor.

Since I started this, I'll download a copy of CWSkimmer and see if I'm right or wrong (will admit it but I know I'm not hi hi ).  People give computer far too much credit - PSK-31 is the result of clever humans devising a method that can be easily decoded - CW predates such decoding plans hi hi.  Computers are still far behind humans in many ways.

Cheers and 73,
Dan
N6ERD

----- Original Message ----
From: Art W6KY <w6ky at yahoo.com>
To: Daniel Severance <daniel_severance at yahoo.com>; Art W6KY <W6KY at Contesting.Com>; n6tj at sbcglobal.net; "Nccc at Contesting. Com" <nccc at contesting.com>; sccc at contesting.com
Sent: Tuesday, April 29, 2008 1:22:31 PM
Subject: Re: [SCCC] Anti-Skimmer Petition

Sorry,  PSK-31 has no error correction...More sophisticated modes
of PSK do but are rarely used on th ham bands...
Decoding 175 pieces of information on CW per minute at 35 WPM
vs aprox 1700 (35 wpm PSK) pieces of digital PSK information
would be much easier, especially in a noisy environment. Remember,
CW is digital....
This program, (like N1MM RTTY mode) can use the Master.dat
file and highlight 'valid' vs 'invalid' contest calls...
Time to buy a SOB (Shack On the Belt) and start talking thru
repeaters about politics and hemorrhoid operations!

73, Art  W6KY
www.w6ky.com

----- Original Message ----
From: Daniel Severance <daniel_severance at yahoo.com>
To: Art W6KY <W6KY at Contesting.Com>; n6tj at sbcglobal.net; "Nccc at Contesting. Com" <nccc at contesting.com>; sccc at contesting.com
Sent: Tuesday, April 29, 2008 1:01:51 PM
Subject: Re: [SCCC] Anti-Skimmer Petition

Hi Art,

PSK-xx has error correction built into the signal to help with the decoding, thus, the ability to decode signals you don't even hear.
CW has no set of checksums, error bits or anything imbedded - it's just dots and dashes.  More than half the time when I use a reader to see my call come back, it comes back as i6erd (missed the first dash) but I hear it correctly and send my response.

CW was not designed as a noisy enviroment digital mode like PSK-31, or even some of the more sophisticated modes with even more error correction built into the mode.  Any click, pop, etc. shows up in the signal and there is no way for the software to unambiguously figure it out without programming in all of the complex  thought patterns that you use when you listen.  You hear xx5xxx but that doesn't make sense given where your beam is pointed and on second thought you figure out it must be yx5xxx, and a quick check in super check partials shows that to be the most likely.  You come back to him and bingo, you were right.   You do a LOT of processing along with hearing the weak signals that is FAR too complex for the computer to do (without a NASA sized budget).

73
N6ERD
Dan

----- Original Message ----
From: Art W6KY <w6ky at yahoo.com>
To: Daniel Severance <daniel_severance at yahoo.com>; n6tj at sbcglobal.net; "Nccc at Contesting. Com" <nccc at contesting.com>; sccc at contesting.com
Sent: Tuesday, April 29, 2008 12:35:01 PM
Subject: Re: [SCCC] Anti-Skimmer Petition

Weak to unreadable signals will be decoded. Go to 14070.15. No signal can be
detected by your ear. Fire up 'Digipan' and readable signals you can't hear appear
on the screen.. This program is decoding over 30 signals at a time in 3kHz in 
PSK-31. It will even 'highlight' a signal calling CQ or QRZ...
So how easy is it for a skimmer program to decode dots and dashes vs PSK...
Also with the VE3NEA Skimmer program, all callsigns are extracted and posted
to the international cluster. Yowiee!
Also, who's to say, if this catches on that Yaesu ans ICOM won't have that
additional wideband output like the SDR and Flex radios...
This program has the ears of a 1958 Novice on steroids with an S53A 'wideband'
receiver..Check out the screenshots!
73, Art   W6KY

----- Original Message ----
From: Daniel Severance <daniel_severance at yahoo.com>
To: n6tj at sbcglobal.net; "Nccc at Contesting. Com" <nccc at contesting.com>; sccc at contesting.com
Sent: Tuesday, April 29, 2008 11:46:54 AM
Subject: Re: [SCCC] Anti-Skimmer Petition

Hi,
    I'm a newbie in about every way possible, but I took the time to read the petition and about the only partially valid point I see is in the first paragraph:

> We feel that Skimmer will substantially change the CW Contesting
activity enabling operators to 
> simultaneously copy every signal on the
band - a feat dozens of operators could not do without this technology. 

That's what it does, but there is still a  caveat, since weak calls are unlikely to be properly decoded by the skimmer, while a human is still more capable of picking out and decoding these weak and wavering signals, so advantage human on that weak mult that the pure skimmer op doesn't go after.  It's not a perfect piece of technology.

The next paragraph:

> A dramatic increase in confusion and QRM is expected as pirates will
call CQ to see the callsign they used 
> on a Skimmer list. Hundreds of stations will occupy frequencies in an attempt to run stations when
their 
> signal is not strong enough to hold a frequency, etc.

The spots are local, not broadcast, so the Pirate comment is irrelevant - it's not a packet network.  The comment on weak stations is also irrelevant - you call CQ, find no one coming back, and figure out it's time for S&P - nothing about a skimmer will change that given my comment in the previous paragraph - weak little  signals won't be decoded.

The last paragraph (skipping the short redundant one):

> Further, the next logical step will be automated QSO Machines which,
over time, 
> will actually replace the CW operator in Contesting. 
> RTTY is a mode of operation that already requires a computer to copy
what is being sent. 
> We feel that if Skimmer type technology were
allowed in this mode of contesting automated 
> QSO machines could be
developed and an appropriate category could be designated so 
> that this
technology could be enhanced and those who enjoy that type of automated 
> operation would have the ability to use it. 

Last I checked, no one has developed automated RTTY and/or PSK-31 machines, which are designed for computer decoding - again the strong signals are easy no matter how you do them - weak ones require and op to sample around the signal to find the spot where the weak  signal is best decoded  (3B8 for instance).   The chances of getting caught with a contesting robot is high since many people deviate in responses due to inexperience or personal preference, and the robot is going to give inappropriate responses (unless SAIC, NASA, or some other huge budget org develops it) which can be reported by numerous ops or observers (using skimmers!).

I'm a computer guy - computers are really pretty dumb - a lot of things COULD be done but it's incredibly difficult to make it bullet-proof and there just isn't enough financial benefit to make it worthwhile.  Anything done badly will show up quickly (much easlier to spot than using packet cluster).

Again, I'm a new ham, new contester, new everything EXCEPT computer N6ERD hi hi - so the only question I see is whether you treat it like an enhanced packet cluster (tailered to your QTH by direct observation) and call it assisted, or just let people use it  and suffer if they rely on it completely (as they will miss a lot)...

It will NOT help much fishing calls out of a pileup (only when the background dies and a strong signal comes in), it's almost strictly a S&P aid.

Just my $0.02 and 73,
N6ERD
Dan

----- Original Message ----
From: "n6tj at sbcglobal.net" <n6tj at sbcglobal.net>
To: "Nccc at Contesting. Com" <nccc at contesting.com>; sccc at contesting.com
Sent: Monday, April 28, 2008 11:16:04 PM
Subject: [SCCC] Anti-Skimmer Petition

Fellow Members.

If you also think that Skimmers contribute nothing to CW Contesting, I'd
like you to consider  signing the petition K5GO has started.  See the link
below.

Vy 73

Jim Neiger   N6TJ

> Hi,
>
> I wanted to draw your attention to this important petition that I recently
> signed:
>
> "CW Skimmer and Related Enhancements"
> http://www.ipetitions.com/petition/skimmer?e
>
> I really think this is an important cause, and I'd like to encourage you
> to add your signature, too. It's free and takes less than a minute of your
> time.
>
> Thanks!
> ---------------------------------------------
> 

_______________________________________________
SCCC mailing list
SCCC at contesting.com
http://lists.contesting.com/mailman/listinfo/sccc
_______________________________________________
SCCC mailing list
SCCC at contesting.com
http://lists.contesting.com/mailman/listinfo/sccc