fft of usrp signal

[Giuseppe Ferrante]

Hi, we are three students of UNICAL (University of Calabria - Italy)!
We are trying to receve a GSM signal from a USRP v.1. Our target is
get this signal and discover the transmission frequencies of the
closest BTS. so we are using simulink connecting the ursp_source to
spectrum_scope block of simulink. the spectrum scope block show us the
fft between -128 /+128 KHz of cenral frequency, but is not enough for
us, because we need all the signal from -1/+1 Mhz between the central
frequency. our problem is how to set the spectrum scope parameter in
order to get the signal fft between -1/+1 Mhz!
could you help us? i hope that you understood the question! our
english is not very good :)

[Nagel]

Guiseppe,
the ADCs on the USRP sample a signal at 64 MSPS that means, if you
wanna receive a bandwidth of 2 MHz (-1/+1MHz) you have to configure
the decimation parameter with:

64e6/2e6 = 32

Your next problem will be probably that your pc is not fast enough to
handle all these data and you'll get overruns. I would recommend to
store a few seconds into a file and then work with these data offline.

USRP_sink --> To Workspace/To Mat File

From Workspace --> Spectrum Scope

Hope that helped

Stefan

[Michael Schwall]

Hi,

if you want to increase your observable bandwidth, you have to change
the decimation factor in the usrp-source block. the ad-converter works
with a sample rate of 64 MSPS. To get a bandwidth of -1/+1 MHz = 2 MHz
you have to set the decimation factor to 32. But, the lower the factor,
the more real-time problems (over- and under-runs uU oO) you may get.

Greetings

Michael

--
Karlsruhe Institute of Technology (KIT)
Communications Engineering Lab (CEL)

Dipl.-Ing. Michael Schwall
Research Associate

Kaiserstrasse 12
Building 05.01
76131 Karlsruhe, Germany

Phone: +49 721 608-6276
Fax: +49 721 608-6071
Email: michael...@kit.edu
Web: http://www.cel.kit.edu

KIT – University of the State of Baden-Wuerttemberg and
National Laboratory of the Helmholtz Association

[elizabe...@gmail.com]

Hello all,

 

I am really experiencing problems in using the ifft and fft Simulink blocks in my OFDM model. After processing the signals and passing them through the QPSK block, I am able to retrieve the data back at the receivier end, using USRP Tx and Rx. However, when I try to include ifft block after the QPSK modulator at the Transmitting end, and as well include the fft block before the QPSK demodulator at the receiving end, the output message gets corrupted. I have attached a model to this message, so that you could understand the scenario well. Kindly help me to fix this problem. What am I misunderstanding?

 

Nb: I also tried to normalize the signal, by multiply the transmitted OFDM frame with sqrt (leght) and dividing it by the equivalence at the receiver. Kindly have a look at the model. After running them, you can try to remove the ifft + gain blocks (at the Tx) and remove the fft + gain blocks (at the Rx).  This results into a pure QPSK Tx and Rx. with This, the models work perfectly. So, what is the problem with ifft and fft?

 

Liz 

[Gayathri Kongara]

Hi Liz,

 

I am trying to answer your question based on my previous experience with the use of Simulink and USRP. As far as I can remember, FFT/IFFT operations are carried out in block lengths of 2^n.Hence, in implementation, it requires a databuffer block before the use of FFT/IFFT blocks. I cannot directly see from the Simulink models if you are buffering your data anywhere. In addition, the chosen FFT block length should match the USRP source and Sink processing lengths.

 

All the best.

 

Gayathri

--
You received this message because you are subscribed to the Google Groups "Simulink-USRP" group.
To unsubscribe from this group and stop receiving emails from it, send an email to simulink-usr...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.