Hello Andy and group,
you wrote:
> What I don't know about Wolf, are any details about how the decoding
is done. Although Opera isn't documented, it is quite obvious how the
decoding must work to acheive the performance it does, and is no doubt
very similar to that in WSPR/JTxx which is well documented But the
overlying of repeated messages in Wolf is in a different area, and
perhaps its author could say a bit more.
<
It would be nice if Stewart Nelson, KK7KA was still around, but I
haven't heard from him for a long time.
So here's my two cents about the decoder. I didn't change it for the
WOLF GUI; the algorithms in the following sourcecode archive is
basically unchanged:
http://www.qsl.net/dl4yhf/wolf/wolf_source.zip
The decoder is, like the rest of the 'core', in wolf.cpp .
It includes the pseudo-noise correlation, Viterbi decoder, and all the
'interesting' stuff.
The red tape (soundcard audio handling, using an extra worker thread) is
in WolfThd.cpp.
The WOLF decoder runs in it's own thread. WOLF_Main() called from
WolfRxThdFunc() is a replacement for the old main() in the DOS program.
It repeatedly calls rcv() for reception. rcv() then calls a bunch of
other subroutines - I'd have to examine the code myself to understand
what each of these functions does. To be honest, there are a few
functions which I don't understand myself completely. So just a few
points: The decoder doesn't correlate along the time scale, but also
tries different frequencies (which is the reason that the decoder
requires more CPU power depending on the frequency range).
A short description of the files is in "readme_1st.txt", but since the
innards of the decoder and encoder have not been modified (compared with
Stewart's original work), I didn't take the time back then to describe
how the decoder works exactly.
Have a nice sunday,
Wolf DL4YHF .
In fact, I'd say Wolf has beenhiding its possibilities in the bushes,
capable of far greater development than non coherent modes and probably
even more effective on the LF bands than OP32. Don't forget, the first
Transatlantic reception of 73kHz used Wolf. Back in the last century
Andy www.g4jnt.com 2012/1/28 Stefan Schäfer
<[email protected]>:
...so Andy, would you suggest it may be useful even on VLF, where the phase
is rather stable over a longer time?
73, Stefan
Am 28.01.2012 20:18, schrieb Andy Talbot:
|
