Hi, Stefan I have tried patch file you gave me, but the result was even worse. The receiver cannot synchronize with the transmitter.
I also tried the DBPSK and DQPSK scheme without your patch file. I found there is a big problem with USRP. After having TX and RX rest for 20 mins, the communication between them are so good that the PDR detected at the RX is almost 99%, but after those two USRPs run for a while, say 10 mins, the communication quality deteriorates rapidly, which means the synchronization was worsen. So, the conclusion could be reached that USRP could not keep on running for a long time, even 10 mins' running would exhaust our USRP? I do not know whether my conclusion is right or not, but if I were wrong, please pick out my problem! Thanks a lot! Bill --- On Thu, 3/19/09, Stefan Brüns <[email protected]> wrote: From: Stefan Brüns <[email protected]> Subject: Re: [Discuss-gnuradio] some results about D8PSK (problems) To: [email protected] Date: Thursday, March 19, 2009, 11:18 AM Hi Bill, I mentioned some problems with carrier offsets between different USRPs (or daughterboards, for RFX non-MIMO boards) when you reported on DBPSK. I found that error rate vastly improved when I factored the offset directly into the frequency arguments of the benchmark_{rx,tx}.py programs. Although the costas loop should be able to deal with carrier offsets, as long as these stay in the allowed range, there is a bug in the mpsk receiver block, which leads to bad performance for larger offsets. Can you try the attached patch? (Lacking time to test it myself atm ...) Stefan On Wednesday 18 March 2009 22:19:50 Bill Stevenson wrote: > ________________________________ > From: Bill Stevenson <[email protected]> > To: Bill Stevenson <[email protected]> > Sent: Wednesday, March 18, 2009 5:16:26 PM > Subject: Re: [Discuss-gnuradio] some results about D8PSK (problems) > > > Hello, > > I have searched the mailing list for the questions about D8PSK and also > reviewed the README file for the benchmark_tx.py and benchmark_rx.py, it > seemed that D8PSK could not work in a decent way like DQPSK or DBPSK on > USRP now. Is that right? I have tried to find out the most appropriate > values for --costas-loop and --gain-mu, but still found that 0.05 for the > gain of loop filter and 0..001 for the gaim-mu could make the PDR reach > 100% with two USRPs seperated with very short distance. Other values for > these two arguments would ruin the communication seriously. So the D8PSK > could not work well on USRP??? > > Thanks. > > Bill -- Stefan Brüns / Bergstraße 21 / 52062 Aachen mailto:lurch at gmx.li http://www.kawo1.rwth-aachen..de/%7Elurchi/ phone: +49 241 53809034 mobile: +49 151 50412019 -----Inline Attachment Follows----- _______________________________________________ Discuss-gnuradio mailing list [email protected] http://lists.gnu.org/mailman/listinfo/discuss-gnuradio
_______________________________________________ Discuss-gnuradio mailing list [email protected] http://lists.gnu.org/mailman/listinfo/discuss-gnuradio
