[CQ-Contest] Rule Change Debate on Skimmer

[Lee Volante]

It's been very interesting to watch the latest round of the 'Skimmer debate' 
over the last couple of weeks.  It has made us realise we have some 
different opinions of what  activities should be classed as valid for 
'single operator unassisted' - and some of these activities have been around 
for far longer than Skimmer itself.  Below are a few of my thoughts for the 
melting pot.



There have been 'no code' hams taking part in CW contests for quite a while, 
and I'd guess that number has increased worldwide in the last couple of 
years with the international relaxation of morse code proficiency for access 
to the HF bands. We welcome them taking part in contests however they feel 
able.  Some will then learn the code, (and we thank them doubly for this), 
but some won't. It's a fact of life.  Some who are learning will be using 
some software or hardware as a crutch whilst they gain proficiency.



Let's consider the case of a ham with a code reader who enters a contest. I 
think until about 3 weeks ago nobody cared that much if he entered in the 
Unassisted category if it was just him and his code reader. There wasn't 
much fanfare on this reflector at least. The technology used in 99% of cases 
was just decoding one signal at once, and with an efficiency and accuracy 
less than that of an average 'human op.'  So, there was no risk of upsetting 
the apple cart with machine triumphing over man. As our new contester 
develops - some of the older hands might prefer that his entry with a code 
reader does not by default place him in an Assisted category.  If he was, 
there would be a feeling of compulsion to use the DXCluster to compete in 
his category, and the potential risk that the skill of searching for QSOs 
and mults all by himself be lost on him.



Some PSK software has been able to decode multiple steams in the audio 
passband for a time already, using a standard PC and ordinary radio. When I 
first saw this, I realised the advantage this gave over my single stream 
setup.  The better software gave simultaneous decoding of callsigns 
everywhere in the PSK sub-band, and you'd know exactly when it was time to 
call another station. This is the 'Skimmer + SDR nightmare' some folk have 
been predicting, but already realised. If PSK contesting was bigger than it 
was, I guess the present debate would have been thrashed out already.



At the present CW, SSB and RTTY contests have pretty much the same 
definition of unassisted and assisted categories. This is especially 
important for those mixed mode events in the calendar. From some of the 
recent comments made here, all RTTY operation is by default 'assisted' and 
new classifications are necessary. This would upset the symmetry we have at 
the moment and make comparisons with historic results more difficult.



My own two-cents worth suggestion is that the line drawn between 'unassisted' 
and 'assisted' be based on bandwidth.  If an MFJ code reader or Skimmer is 
helping to decode a single signal on one frequency, it could be classed as 
unassisted. No assistance is being given decoding callsigns on other 
frequencies, in the same way as DXCluster or a second operator would (which 
we are all happy would place the operator in an 'assisted' or 
'multi-operator' category.)  This setup should not outperform a human 
operator.  If the decision to transmit remains with the human being - this 
does not allow an automatic operation which would have greater stamina than 
a human over long events.



This approach would be compatible with RTTY and PSK contests, and maintain 
the status qso.  When MTTY decodes RTTY callsigns for me one at a time on 
screen on my run frequency, I'd be classed as an 'unassisted' entry as I 
have been for years.  If the scope of the "detection", for RTTY (or for CW, 
or SSB) grows in bandwidth to spread up the band, then it would fall into 
the "assisted" category.  That would be achieved via an SDR wide bandwith 
receiver, or the automatic tuning up and down the band via software.  More 
information is presented to the operator than could be gathered by himself 
in the same time. The extra information gathered is very much akin to that 
of the traditional DXCluster.  For SO2R we'd perhaps need to allow 2 'spot' 
frequencies for an unassisted operation.



Some people obviously have very different views and ideas, and have already 
aired them, but I just wished to share mine.



73,



Lee G0MTN