You are picky AND correct! Andy.
On Dec 6, 2007 10:33 PM, Rud Merriam <[EMAIL PROTECTED]> wrote: > > > > > > > I think I am correct, and admittedly picky, to say that what is described is > a form of source encoding with QPSK modulation. > > QPSK only applies to the modulation of two independent bit streams in a > signal. A clear way of looking at this is that based on I/Q modulation the > I > carries one bit stream while the Q carries the other. > > The encoding below determine how the two bit streams are created. > > The URL did not work for me. > > Rud Merriam K5RUD > ARES AEC Montgomery County, TX > http://TheHamNetwork.net > > > > -----Original Message----- > From: digitalradio@yahoogroups.com [mailto:[EMAIL PROTECTED] On > Behalf Of Andrew O'Brien > Sent: Thursday, December 06, 2007 9:19 PM > To: digitalradio@yahoogroups.com > Subject: Re: [digitalradio] Re: Technical Question: FDMDV / QPSK in PSK31 > > On Dec 6, 2007 9:39 PM, Vojtech Bubnik <[EMAIL PROTECTED]> wrote: > > > > > > > It is QPSK, not BPSK. > > > > 73, Vojtech > > > > > > > > And while on the subject of QPSK, here is an old article about PSK31 > QPSK that I wrote around 2000 for the Logger32 help files, it > includes quotes from Peter Martinez that he granted permission for me to > use > . > > Andy K3UK > > QPSK (Quaternary Phase Shift Keying) Operation > Andrew J. O'Brien, KB2EOQ > > QPSK is referred to as an error-correcting PSK mode. Strictly speaking, it > is not error-correcting in the traditional sense, but we'll leave that for > another discussion. You will find it very useful in copying very weak > signals. You will also discover that it is used less than BPSK. The usual > convention among QPSK'ers is to use BPSK to establish a QSO and then switch > to QPSK You will rarely find someone calling CQ in QPSK mode. Some radio > amateurs will switch to QPSK if a BPSK QSO with weak signals is producing > poor copy. PSK veterans will point out, however, that while QPSK > outperforms > BPSK when weak signals are the issue, QPSK will perform no better than BPSK > if noise is the signal limiting condition. In the ancient, early days of > PSK31, in the last century, QPSK was used as the mode of choice for some > PSK > nets, but that does appear to be the case in the 21st century. . Tuning a > QPSK Signal snip.... > > QPSK can perform better than BPSK when band conditions are poor due to > polar > flutter. At such times you may be able to get copy from a QPSK signal even > though the "cross" in the tuning window is poorly formed. > Note: QPSK requires that BOTH stations in the QSO be using the same > sideband! While this is not important for BPSK QSOs, it vital when using > QPSK. > > QPSK-Reversed > BPSK is demodulated the same way whether you are on upper or lower > sideband. > This is not true for QPSK, which operates like RTTY. You must shift the > signal in a way that the decoder expects or it will not decode. In the case > of QPSK, this leads to problems with standards, since the mode is so new. > In > actual practice, most hams appear to be operating BPSK using AFSK and upper > sideband. This means that, when they switch QPSK (without reversing), they > must be decoded by the other station using upper sideband and QPSK (without > reversing). However, the standard of RTTY operation is lower sideband, and > this means that most hams, operating as just mentioned, are operating > QPSK-reverse, based on the RTTY standard. What this really means is that, > when you and another station decide to switch to QPSK, if you do not know > whether that station is using upper or lower sideband, there is a chance > that you will not decode that station. If you click again on the mode pane, > it will switch from QPSK to QPSK-reverse, and you should begin to copy. One > trick is to set all three Rx windows on the signal to be copied, set one > aux > window at QPSK and the other at QPSK-reverse. When you see which Aux window > starts to print readable copy, switch your main Rx window to that mode and > you can then > > So What Is QPSK? How QPSK Got Its Name > Peter, G3PLX, says in an article entitled "PSK-31, A new radio-teletype > mode > with a traditional philosophy," says that he called it "quadrature polarity > reversal keying" (which of course would have come out as QPRK), but that > everyone else calls it quaternary phase-shift keying (QPSK). It is an > error-correcting mode that relies on four phase-shifts rather than two, to > create the basic scheme. > > Peter Martinez' (G3PLX) Description of QPSK > The QPSK mode used in PSK31 takes the binary data-stream, at the point > where > it would otherwise go direct to the BPSK modulator, and feeds that through > a > 5-bit shift-register. A logic operation then forms the parity function of > the 1st, 2nd, and 5th stages, and another forms the parity function of the > 1st, 3rd, and 5th, giving two bit-streams at the same 31.25 bps rate as the > original data. These two bits form a binary number, the four values of > which > are mapped to the four possible phase-shifts in the QPSK modulation. Thus a > single data-bit from the source results in a 5-bit-long predictable > sequence > of 90-degree and 180 phase-shifts, interleaved with those of following and > preceding bits. In the receiver, a Viterbi decoder is used to keep track of > all 32 possible combinations of guesses at the transmitted datastream, a > running total of how well the received pattern of phase-shifts matches each > guess. The clever thing about the Viterbi decoder is that it can be sure > that if it throws out the worst 16 guesses before it receives the next > signal (which doubles the number of to 32 again), it can never get it > wrong, > and after about 20 more signals, it can be pretty certain that its guess 20 > signals ago was right. The snag is that it can only output its best guess > after a delay of 20 signals, or 640mS in the case of PSK31. Longer Viterbi > decoders are possible, but the amount of computation doubles at each extra > signal delay, and for PSK31, which is used for live QSOs, the delay would > be > too long. More Facts to Consider About QPSK The relationship between BPSK > and QPSK is that both have the same bandwidth, but that QPSK uses that > bandwidth for two signals, shifted-from each other by 90 degrees. The extra > signal is used to transmit redundant information for error-correction. > reduces the signal-to-noise ratio for QPSK by 3 dB, in comparison to BPSK. > The expectation is that the error-correction will more than make up for > this > difference. The results are not in, and one goal of current PSK31 activity > is to evaluate and compare BPSK QPSK, to determine the situations in which > each method is superior. According to Peter, BPSK should be superior under > conditions of ordinary noise (white noise or random noise) but that under > conditions of fading or real-life interference (as contrasted with random > white noise), QPSK may be superior. More evaluation is needed. For a good > introduction see NB6Z's excellent discussion of QPSK at > http://www.teleport.com/~nb6z/psk31.htm. > *Peter Martinez has granted Logger32 permission to use his above > description > > Announce your digital presence via our Interactive Sked Page at > http://www.obriensweb.com/drsked/drsked.php > > View the DRCC numbers database at > http://groups.yahoo.com/group/digitalradio/database > > Yahoo! Groups Links > > -- Andy K3UK www.obriensweb.com (QSL via N2RJ)
