Sending and receiving (LTE) data through Agilent equipment

[exp]

Hi,

As a simple test, I would like to send and receive a simple signal in MATLAB (!) with a VSG (Agilent N5182A) and VSA (Agilent N9010A).

I did the following steps:

1.) Created a simple LTE signal (20MHz at a certain carrier) with Agilent Signal Studio, exported to MATLAB. The mat file contains a 307200 point vector "b" sampled at 30.72MHz
2.) Connected VSA/VGA, uploaded to VSG
3.) Captured data from VSA with Agilent VSA software and exported to MATLAB. This file does not contain a sampling rate information but an Y vector with the time-domain data, XUnit as 'Sec' and XDelta as '3.125e-8'. From this I derived that the sampling rate must be 1/XDelta=32MHz (!)
4.) Center both vectors by mean and normalize by std. Use cross correlation to match the fragments. The (small) received fragment can not be found in the sent data

Observations:
* When I plot both FFT data on top in MATLAB, they are at least aligned. So I think fs1=30.72M and fs2=32M is correct
* When I use the Agilent tool and set Receive Mode "LTE" is displays me a constellation diagram with the received data. So it should work at least
* Why 32M ??? That's not even an integer multiple of the sent data. Can I change this sampling rate somehow? I did not find out how ...

I assume that the issue is on the way from VSA -> MATLAB export.

Any idea what could go wrong here?

 

[jossion]

Hi,

Kindly set the sampling rate that you need in Agilent studio as well as the VSA. That's the only way out. Once captured if you want to change then the data doesn't make any sense

 

[exp]

Thanks.

However, I am not able to change either sampling rate.

I wonder that's true for >100k equipment but it seems so.
Do you have a pointer where to set this?

For the VSG, I can just sent the generated signal to it. And I cannot change this setting. It's the ~20MHz LTE with a predefined oversampling, resulting in 30.72M.

On the VSA side, I cannot see ANY settings related to that! And generally it should work because the Agilent application itself displays the phase diagram, shows EVM etc.

 

[jossion]

Now as is understand is your problem with the data set after importing to matlab?

 

[exp]

Actually I don't know where the problem is.
Fact is:
* If I use the predefined Agilent application, it shows the constellation diagram and outputs EVM. But I have no idea about sampling rate and other parameters
* When I export the data (or capture) I can not define the sampling rate. It is always fixed at something I don't want (32M)
* When I resample both signals to have the same rate in MATLAB ("resample"), the data I received is not the data I transmitted (using cross correlation and after normalization).

In my opinion, resample should not cause the problem, but maybe it does? As said, the spectra "look" the same but the data is different

A fundamental question is for sure how I can set the sampling rate for an Agilent VSA ...

 

[jossion]

since it is showing the constellation and output EVM I assume you are using the LTE module in Agilent VSA. In that case the sampling will be set according to the standard which they may not be giving. But i thing while exporting the sampling rate may be mentioned somewhere in the file. You may have to read the documentation for understanding that.