[Moon] Deep Search IS cheating

Leif Asbrink leif at sm5bsz.com
Fri Aug 22 23:05:40 CEST 2014


Hi All,

I do not read many of the postings in this discussion because
mostly the technical level is very low - or non-existent.

There is however one point that I think needs to be considered
when the CW cheat-sheet is compared to deep search.

With the limited lists of today, knowing 3 characters might be enough
to pick a station from the CW cheat sheet. I suspect people do not
really do that because once one has found a candidate one can
listen a little longer and then the sensitivity is much better
because one "knows" what to expect. Based on that one has "heard"
the full call sign.

To understand the fundamental difference, assume that the cheat sheet
as well as the call3.txt database contains all the 2(?) million
call signs issued worldwide. JT65 would not be affected, but 
the cheat sheet for CW would be totally useless.

The deep search in JT65 would not be affected. The more different
call signs that are compared to the received data the better.
Ideally the received data should be compared to all possible
messages. The matches should form something like a Gaussian
distribution. If there is one single message that shows a 
significantly better match than any other message and provides a 
data point that is unlikely to belong to the Gaussian distribution
a detect made.

When the call3.txt list is limited like it is today, the number of 
points on the Gaussian distribution is not very large. That may
add some uncertainty in the exact form of the tail of the curve
and make it a bit uncertain how far out a match has to be to be '
considered a detect.

In the past, maybe 7 years ago and earlier, detects were based on
the absolute value of the match. That caused false detects sometimes
because the entire Gaussian may shift a little with different
received signals (when they contain interference that is not white 
noise.) I do not remember any details, but I am sure Joe could tell
if someone would be interested.

I personally think that neither call3.txt nor the CW "cheat sheet"
is cheating. I consider both as legitimate tools to improve sensitivity.

I do however consider the usage of scheds as cheating. Both in CW
and in JT65. I have solid experience of cheating in CW before the
JT era. I was the only station with full polarization control.
I always heard all stations having a single yagi and 500W or more.
There was a sched list on the Internet every month and I used it
to find stations I never worked before. (Newcomers who only
used scheds.) I would listen and after maybe 2 periods I would
know that the sched partner is QRT. That happened now and then.
I would call the interesting station KH6ABC de SM5BSZ SM5BSZ 
SM5BSZ. The result was always (except once with a VK station)
that the interesting station would send the call of his sched partner
and then OOO OOO OOO. I would continue sending mostly my own call
and the other station would continue OOO to his sched partner.
After 10 minutes I would quit.

It is OBVIOUS that most sched QSOs are based on "someone is on
the frequency so it must be my sched partner" I know for
sure that I produced a strong signal with the correct polarization.
The sched station just did not care to listen (or did not know CW)

I have done this test about 10 times and 1 in 10 for not cheating
is a very discouraging result. I actually think the integrity 
of JT65 contacts is better provided that there is no parallel
Internet communication.

I think, contrary to use lists, that basing the identification
on the presence of a signal on the correct frequency is cheating.

Random contacts are fine CW as well as digital as I see it.
That does not mean that I consider a DXCC with JT65 anywhere
as valuable as one with CW only.

The big stations can produce SSB that is easily heard by most
stations. Mixed mode JT65/SSB should be more common. Mixed mode
JT65/CW is of course technically more feasible, but all JT65
operators do not know CW...

Personally I find JT65 boring. The format is too restricted
and it only allows collecting new stations like collecting stamps.

In the CW era, breaking the rules cleverly, was what made
operation exciting!

73

Leif / SM5BSZ



More information about the Moon mailing list