how to get S21 from usrp 2901 using labview 2014

[Tawahaa38]

Hi

I want to get S21 from usrp 2901 using labview 2014. Is there any way I can achieve this goal.

Please do suggest m any alternative as I want to differentiate the transmitted signal from received signal.

Thankss

[Hey.itsDale]

Hello,

The 2901 is not a calibrated device and furthermore it is not designed for s-parameter measurements. However, you might be able to get an approximate value for the magnitude of S21. I'm unsure if getting phase information would be possible.

To do this you could try the following:

1. Transmit a tone on the TX port with a power that will not saturate the input/output of your device
2. Receive the tone on the RX port and record the power as RX_1
3. Insert your DUT between the TX and RX port
4. TX the tone again and record the RX power as RX_2
5. Take the ratio of RX_2/RX_1
6. Repeat for all frequencies of interest.

This should give you an approximate value for S21. However this is not the intended use of the 2901 and there is no way to know how accurate your measurement is without comparing to a calibrated VNA.

For further information, please see this white paper which gives a nice overview of VNAs

http://download.ni.com/evaluation/rf/Introduction_to_Network_Analyzer_Measurements.pdf

[Shalini.M]

Hi Tawahaa38, 

The USRP 2901 is not recommended for s-parameter measurements for 2 reasons:
1) It is not a calibrated device and should be used for prototyping purposes instead of testing

2) The USRP does not have a coupler which is a critical component that VNAs have for measuring the transmitted vs. reflected power. 

From my understanding, it sounds like you should be using a VNA for your application. Can you give me more information on what your application and end goals are? Can you also elaborate on what you mean by differentiating the transmitted signal from the received signal? 

[Tawahaa38]

I am actually making a metal detector using Rf waves. So currently what I am doing is I am transmitting a signal from one 2901 and receiving it on another 2901 using log periodic antenna. Then applying some signal processing techniques to differentiate between transmitted and reflecting signal.

We do have a VNA in our lab but my task is to use usrp for this project. 

So could we use an external coupler with 2901 for getting better S21 or there is any way that VNA can help enhance the results of 2901.
I have also attach my labview files.

Thanks

[Tawahaa38]

I am actually making a metal detector using Rf waves. So currently what I am doing is I am transmitting a signal from one 2901 and receiving it on another 2901 using log periodic antenna. Then applying some signal processing techniques to differentiate between transmitted and reflecting signal.

could you please elaborate in terms of block of codes in labview.
I have also attach my labview files.

Thanks

[Hey.itsDale]

That sounds like an interesting application, but unfortunately I don't have any direct experience with metal detection. Is there a specific algorithm you are interested in implementing? Could you explain what signal processing techniques you are using?

Given an algorithm, I could hopefully offer some advice on how to code it in LabVIEW.

[Tawahaa38]

I have attached my receiver cod in previous comment my method for signal processing is that 

I want to first take inverse fft of the reflected wave than want to do averaging of the inverse fft then then aligned then by correlation coefficient and eventually sorting them by amplitude.

[Hey.itsDale]

Could you elaborate on what the IFFT is for? You will be getting time domain data from the USRP which means you could use an FFT right?

Also, it looks like you are transmitting a sine wave. If you want to discriminate between signal coupled from your TX antenna and signal reflected from a detected object, I'm not sure a sine wave will work. The signals will add to form another sine wave at a difference amplitude and phase.

[Tawahaa38]

i have actually read a resaech thesis for metal detection in that thesis the author was taking ifft of the S21 from VNA that's why I am taking IFFT. If you have a better idea do let me know. Please tell me what would be the good option for the type of transmitting wave as you said sine wave might not work.???

[Hey.itsDale]

Interesting! I was not previously aware that s-parameters could be used like that. What type of signal was used in the research thesis? Perhaps you could describe their hardware setup so we could compare that to the USRP-2901?