"Bevan Weiss" <kaizen__@NOSPAM.hotmail.com> wrote in message
news:Z3C%f.13010$JZ1.521061@news.xtra.co.nz...
>
> If you really mean pi/4-QPSK and you know that pi/4-QPSK is [ pi/4,
> 3/4*pi, 5/4*pi, -pi/4] then what's the problem?
This is incorrect. QPSK uses four signals which can be taken to
be phases
0, pi/2, pi, 3*pi/2,
OR
pi/4, 3*pi/4, 5*pi/4, 7*pi/4.
Note that the latter constellation is just the former rotated by pi/4.
Now, pi/4-QPSK is a modulation scheme that uses the above two
signal sets in alternation; one dibit is sent using 0, pi/2, pi, 3*pi/2,
the next pi/4, 3*pi/4, 5*pi/4, 7*pi/4, the one after that using
0, pi/2, pi, 3*pi/2, again and so on. The advantage over plain vanilla
QPSK in which two successive identical dibits result in no
phase change is that in pi/4-QPSK, the signal phase changes
from dibit to dibit regardless of whether the two dibits are the same
or different. Furthermore, a dibit change from 00 to 11 (or vice versa)
or from 10 to 01 (or vice versa) does not cause the signal
envelope to pass through the origin . This is useful when the signal
has to go through a nonlinearity such as a traveling-wave-tube
amplifier in a satellite.
>
> If you're instead asking about pi/4-Offset QPSK, or pi/4-DQPSK, then
> that's different.
>
> pi/4-OQPSK is simply QPSK where each symbol period the phase is shifted by
> pi/4. So if the same word is transmitted then the constellation point
> transmitted still rotates by pi/4 each period.
This too is incorrect. Offset QPSK (OQPSK) is a modulation scheme in
which the inphase and quadrature bits do not change at the same time. For
a bit rate of 1/T bps, the inphase bit changes at t = ... -3T, -T, T, 3T,
...
while the quadrature bit changes at t = -4T, -2T, 0, 2T, 4T, ... (Some
people do it the other way just to spite us right-thinking folks :-)). This
too
avoids the transition of the signal envelope through the origin. The offset
means the offset in the bit timing in the I and Q branches, not the signal
phase.
> pi/4- DQPSK I believe is similar to pi/4-OQPSK, however it relies on the
> transition between symbols rather than strictly the symbols themselves.
pi/4-DQPSK uses the change in the phase between two signaling
intervals to encode the dibit. Thus, if the new dibit to be transmitted
is 00, the signal phase is ADVANCED by pi/4, if the new dibit is 10, the
signal phase is ADVANCED by 3*pi/4, if the new dibit is 11, the signal
phase is ADVANCED by 5*pi/4, and if the new dibit is 01, the signal
phase is ADVANCED by 7*pi/4.. The phase of the signal prior
to the change can be one of 0, pi/2, pi, 3*pi/2, in which case the new
phase is one of pi/4, 3*pi/4, 5*pi/4, 7*pi/4 (or vice versa) Thus, the
two constellations are used in alternation so that, for example, if the
initial phase is 7*pi/4, then eight dibits 00 00 00 00 00 00 00 00 are
using signal phases 0, pi/4, 2*pi/4, 3*pi/4, 4*pi/4, 5*pi/4, 6*pi/4, 7*pi/4.
--Dilip Sarwate