Re: [Discuss-gnuradio] TETRA demodulator advices

[Sylvain Munaut]

Hi,


> I've looked into that but I'm afraid some blocks won't play well with the
> constellation class (for example, they expect a constellation to have
> 2**(bits) points, but PI/4 DQPSK uses 2 bits and has 8 points).
> I think I will stay with the flowgraph approach for now.

I've worked with pi/4 xQPSK signals a bit when doing tetra and gmr
stuff and the approach I took is just to undo the pi/4 rotation using
a complex rotator.

Then it looks like normal QPSK and can be processed as such.


Cheers,

   Sylvain