| |
PA2AGA > HDDIG 06.07.00 02:06l 166 Lines 7271 Bytes #-9409 (0) @ EU
BID : HD_2000_180B
Read: GUEST
Subj: HamDigitalDigest 2000/180B
Path: DB0AAB<DB0PV<DB0MRW<DB0ERF<DB0BRI<DB0SM<PI8DAZ<PI8GCB<PI8HGL
Sent: 000705/2040Z @:PI8HGL.#ZH1.NLD.EU #:58934 [Den Haag] FBB $:HD_2000_180B
From: PA2AGA@PI8HGL.#ZH1.NLD.EU
To : HDDIG@EU
Date: Wed, 05 Jul 00 21:28:02 MET
Message-Id: <hd_2000_180B>
From: pa2aga@pe1mvx.ampr.org
To: hd_broadcast@pa2aga.ampr.org
X-BBS-Msg-Type: B
>first method of FEC to their transmissions.
>
>In fact many commercial CW operators and certainly most military CW
>operators use FEC. This usage of FEC over CW began in WWII, "the big one"
>and has been in use continuously since then.
>
>Of course you won't have heard it on the air in "plain language", but I can
>send you to several HF frequencies where you may copy on air CW FEC,
>generally in the format of "five letter code groups".
>
>In fact, unlike for amateur radio examinations where the testing for CW
>prowress is done in plain language, for commercial and military sending and
>receiving CW tests, the tests invariably include/included code groups, many
>of which are/were FEC encoded.
Ok, I'm interested. What FEC algorithm is used with these code groups?
AFAIK code groups are only used to obscure meaning, and don't include
cross products and other redundancies designed to locate and correct
transmission symbol errors as part of their design. Note that I don't doubt
that such error detecting and correcting block coding *could* be done for
CW transmissions, but I've never heard of it actually being done in practice.
Gary
Gary Coffman KE4ZV | You make it |mail to ke4zv@bellsouth.net
534 Shannon Way | We break it |
Lawrenceville, GA | Guaranteed |
------------------------------
Date: Tue, 04 Jul 2000 12:06:33 -0400
From: Gary Coffman <ke4zv@bellsouth.net>
Subject: CW versus hi speed digital etc.
On Sun, 2 Jul 2000 17:33:53 -0500, "Peter O. Brackett" <ab4bc@ix.netcom.com>
wrote:
>There are some aspects of analog in hand sent CW, this would be interperted
>only as a timing "jitter" effect which all digital communications systems
>from SONET lightwave systems to ELF submarine data systems, have to various
>extents.
>
>In fact even this aspect of CW has been addressed by coherent CW enthusiasts
>who have addressed the timing and synchronous detection problems you alluded
>to.
CCW technique, which is different from CW, does use fixed clock timing, though
it doesn't use uniform symbol timing (it can't, because the symbols aren't
uniform).
The information is still encoded in the symbol edges, which are harder to
detect than
the mere presence or absence of signal.
>Also... from a channel coding and modulation viewpoint, CW can easily be
>transmitted using a variety of modulations from FSK to PSK, to whatever. It
>just then makes direct copy of the sound "by ear" less familiar to
>operators, a conversion device to convert back to a keyed tone would suffice
>for human readability.
One of the problems here is the use of the term "CW". Conventionally that
means use of Morse Code by on-off keyed baseband symbols which are
then impressed on a constant carrier by 2T AM modulation. Anything else
is no longer CW. It would be helpful if you are a bit more precise when you
talk about the character encoding, the baseband modulation, or the
transmission modulation so that we can know what you mean.
>Gary despite what you say, CW is truly digital!
No, I disagree. CW encompasses all of the items I listed above.
That is not a digital system. Taken individually, some of the elements
of CW might be considered digital, but taken as a whole CW is not
a digital transmission mode.
>It is so advanced in it's design concepts that it is not even a simple
>binary system, in fact it is a multi-level digital logic code.
>
>It uses a ternerary code alphabet with three symbols, space, dot, and dash
>these three symbols could be assigned the labels, A, B, and C and sent by
>many modulation methods. Up to and including modern trellis coded, lattice
>coded or turbo coded modulations. What???
Actually, Morse is a quintenary code. It has five symbols, the dot, the dash,
the intracharacter space, the intercharacter space, and the word space.
When conveyed using OOK AM (aka CW), the latter 3 symbols aren't voiced,
and have to be inferred against the noise background. That reduces the
probability of correct detection substantially. When Morse coding is used
with other modulation forms, such as FSK, all 5 symbols are voiced, and
the probability of correct decoding increases.
>Even as CW is conventionally encoded and transmitted this ternary channel
>code incorporates a highly efficient source coding algorithm which is
>variable length and comma free. And this was all invented over 150 years
>ago, long before Dr. Shannon was even born and the concepts of source coding
>and channel coding were elucidated in the modern sense. In this respect CW
>is similar to but more efficient than the Varicode that Peter Martinez
>designed for PSK-31...
Varicode is one of the shortcomings of PSK31. It limits the ability to do
context free FEC compared to the usage of a uniform length code. The
fact that the most used letters are sent with the least amount of transmission
symbol information makes the probability of errors higher than for uniform
length codes. It does save transmission time, but at the cost of a higher
probability of uncorrectable errors. Conventional Morse coding suffers
from this same problem, and when used with OOK AM, additionally doesn't
voice some of the symbols at all. At least PSK31 voices all of its
transmission
symbols (one reason it'll work at signal levels where conventional CW
operators can't even detect the presence of a signal at all.)
>Gary, if you cannot make intelligent contributions to this thread, I suggest
>that you go back to the "rec.radio.amateur.antennas" NG, you know more about
>antennas than you do digital communications!
You are incorrect. I've devoted much more study to digital communications
than I have to antennas.
>And besides, my friend, heh, heh you'r spoiling my fun here on this
>thread...
That's what I do best.
Gary
Gary Coffman KE4ZV | You make it |mail to ke4zv@bellsouth.net
534 Shannon Way | We break it |
Lawrenceville, GA | Guaranteed |
------------------------------
Date: Tue, 04 Jul 2000 12:41:56 -0400
From: Gary Coffman <ke4zv@bellsouth.net>
Subject: CW versus hi speed digital etc.
On Sun, 2 Jul 2000 19:24:12 -0500, "Peter O. Brackett" <ab4bc@ix.netcom.com>
wrote:
>Gary:
>
>You are waffling again...
I am neither waffling nor pancaking.
>Good human operators making use of their pscycho-acoustic abilities and with
>appropriate filters and processing in front of them while copying CW can
>usually always beat your vaunted "modern" digital modes over a congested and
>noisy HF channel. [The comparison has to be calibrated at the same data
>rates, of course humans can't operate at thousands of bits per second, but
>that's not the point. The point is that humans can beat the modern digital
>systems using CW at the same information rates! Period!]
That's simply not true. The best known contemporary illustration that it is
not true is PSK31. Time and again operators demonstrate the ability to
work PSK31 signals that are *inaudible* to the ear. PACTOR II also illustrates
the ability of well designed systems to work at signal levels that are as much
To be continued in digest: hd_2000_180C
Read previous mail | Read next mail
| |
