[Discuss-gnuradio] Attempting DBPSK again

[Francesco B.]

I've lost the previous thread amidst other threads, and figured enough new
developments and different issues have arisen to warrant another.

I've written a simple pipeline, which generates a signal, sends it through
the dbpsk modulation block, and transmits it over the USRP. A receiving USRP
intercepts this and transmits anything it receives. Finally, the original
computer is also waiting to receive from the USRP, and stores the results in
a file sink.

The results leave much to be desired, though. Even assuming the most basic
sort of transmission, where all USRP input is monitored and stored rather
than detecting when data is incoming (generating a constant signal by
removing gr.head() from the transmission executable), it still doesn't
function properly. It compiles, and runs, and outputs... but I get a file of
significantly smaller size than what has been transmitted. And even more
curiously, the received file isn't playable by audio_play.py. It runs, but
there is no sound. Even if audio_play.py were given pseudo-random numbers
(supplied by urandom), it would still output sound, even if white noise.

This would all be far simpler if I were aware of how to remove the
modulation aspect from benchmark_tx. The goal here is to have the capability
to supply a modulated bitstream from another source, and have that
transmitted, rather than GNU Radio handling modulation. We would probably
get better results in the end anyway, if we made modifications to a
previously tried method. But... knowing what's wrong here is still quite
helpful.

Code is as follows:
http://www.nabble.com/file/p20964979/loopback_transmitter.py
loopback_transmitter.py 
http://www.nabble.com/file/p20964979/loopback_reciever.py
loopback_reciever.py 
http://www.nabble.com/file/p20964979/loopback.py loopback.py 


-- 
View this message in context: 
http://www.nabble.com/Attempting-DBPSK-again-tp20964979p20964979.html
Sent from the GnuRadio mailing list archive at Nabble.com.