USRP1 can't correctly receive packet from USRP2 using benchmark_tx/rx or benchmark_ofdm_tx/rx

Hi,

I have currently set up a transmission between USRP1 and USRP2 with
FLEX900.
And got the following results.

In gnuradio/gnuradio-example/python/digital/
Successful transmission (ok = True) when
Tx (USRP1) : ./benchmark_tx2.py -f 900M --tx-amplitude=0.4 -m dbpsk2
Rx (USRP2) : ./benchmark_rx2.py -f 900M -m dbpsk2 -e eth0
Fail (ok = False and lost lots of packets) when
Tx (USRP2) : ./benchmark_tx2.py -f 900M --tx-amplitude=0.4 -m dbpsk2
-e
eth0
Rx (USRP1) : ./benchmark_rx2.py -f 900M -m dbpsk2

In addition, if using the original modulation like dbpsk not dbpsk2, the
transmission also not work as expected.

also ref to the two discussions:
http://www.ruby-forum.com/topic/203035
http://www.ruby-forum.com/topic/202105

I successfully transmit packet from USRP1 to USRP2 with the following
parameter setting.
Tx (USRP1) : benchmark_ofdm_new_tx.py -f 900M --tx-amplitude=0.4 -i
128
Rx (USRP2) : benchmakr_ofdm_new_rx.py -f 900M -d 100
But the same problem as above, USRP1 can’t receive correct packet from
USRP2.

Does anyone has any suggection about this problem, please let me known,
Thanks.

Sincerely,
Fisheep


View this message in context:
http://old.nabble.com/USRP1-can't-correctly-receive-packet-from-USRP2-using-benchmark_tx-rx-or-benchmark_ofdm_tx-rx-tp28603392p28603392.html
Sent from the GnuRadio mailing list archive at Nabble.com.

You could be running into the issue of frequency offset between the LO’s
of the two usrp devices. Try increasing the rate so the symbols are
wider spectrally to make up for the offset. I believe that its the -r
option.

Hope that helps,
-Josh

On Wed, May 19, 2010 at 1:48 AM, Josh B. [email protected] wrote:

eth0
parameter setting.
Tx (USRP1) : benchmark_ofdm_new_tx.py -f 900M --tx-amplitude=0.4 -i 128
Rx (USRP2) : benchmakr_ofdm_new_rx.py -f 900M -d 100
But the same problem as above, USRP1 can’t receive correct packet from
USRP2.

Does anyone has any suggection about this problem, please let me known,
Thanks.

Sincerely,
Fisheep

The LO offset would be my first guess, too.

The other issue is that when you specify the bandwidth of the signals,
you set the interpolation and decimation rates. The same inter and
decim on the USRP1 and USRP2 will produce different bandwidth signals,
which won’t be able to talk to each other.

Tom