[Skimmertalk] Skimmer Tests during SS

[Jim Baremore]

I conducted some testing using Skimmer during this years SS contest even
though I did not operate in the contest.  I wrote a rather long paper
describing the details of the test and sent it to Pete, N4ZR, at
contesting.com.  He suggested I do a brief summary and post it here and he
would host the original on their web site for those who wanted to read about
it in more detail. 

 

The hardware for the test was a 7800 serving as a pre-selector for an
SDR-IQ.  A computer ran CW-Skimmer and the output of that was sent to a
Linux Box running some software processing on the data to only pass valid
FCC issued calls to a logging program running on a 3rd computer.  Overall,
for the 13 hours I ran CW Skimmer, The Linux box determined Skimmer copied
12,255 legal FCC call signs of which 2406 were unique.

 

The N3FJP logging program was used to allow me to reaction tune on a spot
and determine if in fact the spot detected by CW Skimmer and verified as a
legitimate US call was in fact the call at the spot location.  Spots were
not worked and I only stayed on the frequency of the spot long enough to do
the call verification and then took the next available spot to test.  The
output of the Linux box, with some additional filters set, was about 500
spots/hour.  I processed spots at a rate of 140 spots/hour.

 

I performed this evaluation at 8 different intervals during the contest and
did evaluations on 20, 40 and 80.  All total 359 spots were tested.  Of
those, only 111 were the station CW Skimmer was indicated on the frequency.
30 were for a station whose callsign was close but which I judged Skimmer
had missed a dit or something and mangled it into another call.
Interestingly enough, the mangled call was still a valid FCC call.  Finally,
218 spots were for stations no longer on the frequency.  Most likely they
were stations just worked by the station running the frequency.

 

The net result is that I was only able to identify new stations in the S&P
mode at a rate of 43 Q's/Hr.  Considering the time it would take to
correctly identify the call and then try to work them would likely make the
Q rate for these spots to be in the 20 - 25 Q/hour region, hardly a
competitive edge.

 

Additionally I note that CW skimmer does not provide section multiplier
information except for a few specialized cases.  That requires an additional
software process, either like I used in this process, or as a master call
data base as part of a logging program as Pete, N4ZR, described in one of
his earlier posts. Without additional software processing, CW Skimmer is not
the same as a Packet/Telnet spotting network that provides a verified call
and section multiplier.

 

My conclusions from the experiment were:

 

I conclude from this experiment that the main value of the CW Skimmer in SS,
without additional software processing, would be to be to assist the
operator in the S&P mode. However as pointed out above, it is not really
providing a competitive edge as one might think given the way the spots
flood the logging program and sorting through that tremendous amount of
information to find a station you can really work is very time consuming.

 

By itself, CW Skimmer provides very limited multiplier information for the
SS contest and thus provides no competitive edge there either. 

 

CW Skimmer is a very significant advance in technology which can quickly
provide a lot of information.  However by itself it is not really the giant
it appears to be at a first look.  It has been put into the same box as
Packet Spotting assistance.  However a PSN not only provides a verified call
and a frequency, it also provides the section multiplier, a very valuable
piece of information.  CW Skimmer does not do that by itself.  I might be so
cynical as to call it a sophisticated band scope.  

 

To be effective, CW Skimmer requires a lot of other software processing of
its' data to figure out who is really running the frequency, what is the
section of the spotted station, and is it really a workable signal from your
location. 

 

Although the conclusion section above is the same as in the write-up,
additional information and my opinions about how CW Skimmer should be
categorized in SS are contained in the original write-up.  

 

73's

 

Jim K5QQ