USRPN210 and Matlab problem in running the QPSK demo

Dear all,

having struggled for a couple of days with some new N210s I decided to
post my problem in here. We recently received two N210s, which I am
willing to interface with Matlab since the drivers have now officially
been released and can be found in:

I have burned the images for the firmware and FPGA on both USRPs. I have
also installed MatlabR2011b on the two computers and configure it so
that it can actually see the devices. I basically have two PCs and two
N210s. Each PC has its own USRP. I attempted to run the QPSK demo TX/RX.
However, I cannot decode the “Hello World” successfully in all cases. In
fact, while the transmitter keeps running, the receiver stops and simply
outputs the decoded message at Matlab’s command line.

Here is what I see at the command line at the RX side, directly after I
click on Simulation–>Start in the sdruqpskrx simulink block:

Opening a USRP2/N-Series device…
---------- end libuhd status message output ----------
---------- begin libuhd status message output ----------
Current recv frame size: 1472 bytes
---------- end libuhd status message output ----------
---------- begin libuhd status message output ----------
Current send frame size: 1472 bytes
---------- end libuhd status message output ----------
---------- begin libuhd warning message output ----------
Unable to set the thread priority. Performance may be negatively
affected.
Please see the general application notes in the manual for instructions.
EnvironmentError: OSError: error in pthread_setschedparam
---------- end libuhd warning message output ----------

I have searched for this error but have not fully understood yet how to
remove it. After compilation and if set Rapid Acceleration mode the
following error comes up, and the simulation stops:

“Unable to find an unused TCP/IP port to connect to the rapid
accelerator target for model ‘sdruqpskrx’”

However, if we set it to Acceleration mode then the simulation proceeds
normally,
but after a small period of time (not greater than 3-4 seconds) the
simulation at the receiver stops and at the command line we get the
decoded word. Below is just a short part of it:

Hello world 052
Hello world 053
Hello world 054
Hello world 055
Hello wU|y 056
S>,hd&/DF g6S
B#WpP0rEW$YLK
ello woHld 059
Hello world 060
Hello world 061

What we would expect however, is a continuous reception and attempt
to decode the message at the RX side, since TX is constantly
transmitting. It
still puzzles me why the receiver cannot keep up at first place
and of course why the message is not decoded successfully in all cases
(distance between the
units is very small). Also the frequency offset is in the order of 1KHz
which can be compensated with this receiver configuration (can deal with
offsets up to 12.5 KHz). Finally, I have varied numerously the gain
parameter at the TX and RX but have not managed to decode the message
successfully at all times.

I apologize for the rather long post but I have made my best to describe
the problem as compactly as possible.

Thanks for your time in reading this.

BR,

Konstantinos

Hi valeris,

I am working with this example at the moment, and I would like to know
what you did to solve the problem. Furthermore, were you able to modify
this example in anyway?