Benchmark_xx.py not consistent output

I am using 3 USRPs (USRP N210 ).
I run example program "benchmark_rx.py " and "benchmark_tx.py " in
/gr-digital/examples/narrowband/.
I performed experiment on 2 USRP’s(A,B) at a time.
so A-B and B-A 2 channels for 2 USRP.
when I send packets from A-B “psk” modulation scheme gives
best(171/171)*
performance while same is not true for B-A(171/0).
while “gfsk” scheme gives good performance from B-A(171/171)
.
Can anybody knows why different modulation scheme gives different result
on different channel ?
Is it related to USRP synchronization ?
Can we perform packet level debugging in USRP ?
How can we find the cause of this problem ?
Or is it implementation related bug in GNU-radio?

  • : (n_rcvd/n_right) n_rcvd => received packet,n_right => correct
    received
    packet

Hi,

I don’t think it is the problem of channel since your channel is
symmetric.
Perhaps you could check your parameters first and list the commands you
used to the mailing list. Usually I use benchmark in
/gr-digital/examples/ofdm and it works well.

BTW, your gmail account is some kind of cool …

Yang, Qing
Information Engineering, CUHK

2012/9/7 usrp n210 [email protected]

On Fri, Sep 7, 2012 at 12:28 AM, Qing Y. [email protected]
wrote:

Information Engineering, CUHK

when I send packets from A-B “psk” modulation scheme gives best(171/171)*
performance while same is not true for B-A(171/0).
while “gfsk” scheme gives good performance from B-A(171/171)
.
Can anybody knows why different modulation scheme gives different result
on different channel ?
Is it related to USRP synchronization ?
Can we perform packet level debugging in USRP ?
How can we find the cause of this problem ?
Or is it implementation related bug in GNU-radio?

  • : (n_rcvd/n_right) n_rcvd => received packet,n_right => correct received
    packet

Try adjusting the frequency to make sure both USRPs are as close as
you can get them by eye. What you are seeing is probably something to
do with the signal being outside the channel and you aren’t able to
bring it back in on one side while the other side is. I’m not entirely
sure why that might happen, but I’ve seen it before.

Tom