Hey folks,
This is a post to the forum, but I'm hoping that maybe the people here can
shed some light on the problem for us.
There is a large misunderstanding between myself and Matt (we're working
on this project together) and everyone else that we ask about PSK
transmission. In a standard transmission and reception of QPSK, your I
channel and Q channel are transmitted in quadrature form using a frequency
of fc. At the receiving end, this signal, I + j*Q, is received, and
demodulated through the use of a standard digital demodulation technique,
such as the one shown here:
[url]http://www.ee.byu.edu/class/ee485/simulink/qpsk/qpsk.html[/url]
The issue that we are having lies in the SDR hardware itself. First,
since the sound card is AC coupled and the data is in digital form, an IF
is used, f_IF. Let's assume our left channel is I_IF and our right
channel is Q_IF. The SDR takes the data from the left and right channels,
I_IF and Q_IF, and upconverts them to RF by using the Tayloe detector. So
at RF, we have one waveform which is a combination of both I_IF and Q_IF.
The sampling clock used in the Tayloe detector is running at f_DDS.
On the receiver, we again have an oscillator controlling the Tayloe
detector. This one is at an arbitrary phase relative to the oscillator on
the transmitter, so we.ll call this frequency f_DDS + phi.
The Tayloe detector samples and holds the RF waveform and produces the IF
waveform to the left and right channels on the sound card. The problem
exists that the sampling boundaries of the Tayloe on the transmitter are
different than the sampling boundaries of the Tayloe on the receiver.
Hence, in one symbol period, the received signal on the left channel is
NOT simply I_IF, it is I_IF for a portion of the symbol period, and it is
also Q_IF for the remainder.
Hopefully the misunderstanding is on our end, and this is a very simple
problem to fix, however we.ve been looking at it for quite some time
scratching our heads, and have been unable to find a mathematical
explanation of the Tayloe detector.
So if anyone can provide a more mathematical explanation of the Tayloe
detector.s operation, or an explanation of our misunderstanding, it would
be GREATLY appreciated.
Thanks for reading all this!
Regards,
Jon
--
*****************************************
Jon Beckwith -KB1KBB- B.S.E.E.
UNH InterOperability Lab R&D
Fast & Gigabit Ethernet Consortiums
121 Technology Drive, Suite 2
Durham, NH 03824-3525 (603) 862-4534
*****************************************
