Re: [Discuss-gnuradio] Re: A little DQPSK demodulation progress and a question

[Eric Blossom]

On Tue, Apr 01, 2003 at 10:34:51AM +0100, Ian Wraith wrote:
> No problem have a look at ..
> 
> http://borg.shef.ac.uk/~ianw/dqpsk_const.html
> 
> This shows 2 graphs one showing a raw data constellation and the other a
> filtered data constellation each showing 500 points of data. This is real
> off air data though as opposed to the usual alternating pattern of 1's
> and 0's which most papers use for their data. That said I am not sure if
> I should be worried by what I am seeing.

If everything was working (including matched filter and symbol timing)
you should see 8 fat dots.  So I think there's a way to go.

To simplify things, you may want to start by building a transmitter
module, so you know for certain what you're sending.  For starters I'd
suggest a sample rate at 4 x the symbol rate.   Forget about the sound
card for for now, run it all in and out of file.  This will allow you
to see the effect of the root raised cosine filter, then you can mess
with the symbol timing etc.  In the simulation, your tx and rx clocks
will be exactly the same frequency, so it removes that variable too.

This will end up being a great intro to digital comms!  

Eric