6G Technology: Wideband Signal Analysis in MATLAB - MATLAB
Video Player is loading.
Current Time 0:00
Duration 3:42
Loaded: 4.45%
Stream Type LIVE
Remaining Time 3:42
 
1x
  • Chapters
  • descriptions off, selected
  • en (Main), selected
    Video length is 3:42

    6G Technology: Wideband Signal Analysis in MATLAB

    Although the 6G standard is not yet publicly deployed, you can still generate and analyze candidate 6G waveforms. With MATLAB®, you can create an 800 MHz pre-6G waveform, transmit it with an NI™ vector signal transceiver (VST), and receive it with an NI X410 USRP™ radio. You can then bring that signal back into MATLAB for error vector magnitude (EVM) analysis.

    You’ll learn about some key capabilities in MATLAB, including wideband signal generation, its ability to connect to software-defined radios and T&M equipment, and its computation of signal quality measurements as specified by 3GPP.

    Published: 15 Oct 2024

    MIKE MCLERNON: Hi there. My name is Mike McLernon. I'm with the Wireless Technical Marketing Group at MathWorks, and I'm here to tell you about a little demo that we have put together on 6G-- not 5G, but 6G. So let me show you a little bit about the hardware.

    First, this large box here, that is an NI VST, Vector Signal Transceiver. So what it has inside, this is an embedded controller, essentially an embedded computer, that is running MATLAB. So also, and it will control this RFSG, RF Signal Generator, and it will come out of this wire with an 800 megahertz wide signal. Now, that far exceeds any max bandwidth that 5G can muster. Max 5G bandwidth is 400 megahertz. We are pumping out 800 right here.

    We go into this mini circuit splitter, and then we go into two channels into this NI X410 USRP radio. 800 megahertz inside, but the max bandwidth on each channel of the X410 is only 400 megahertz. So each channel filters down to 400 megahertz but in contiguous bands. So you've got two 400 megahertz signals. Remember that.

    Then we come back out of the radio, back into the embedded controller and MATLAB, where those two 400-megahertz signals get stitched back together to make the composite 800 megahertz 6G exploration signal, at which point we will do some analysis, some error vector magnitude, EBM analysis on that signal. Here we see some graphical outputs from running these two scripts.

    But before I dive too deeply into that part, I do want to mention the products that we use to create these demos-- MATLAB, of course. But also, on the transmit side to interface with the VST, we use the Instrument Control Toolbox. And then to interface with the USRP radios, that would be Wireless Testbench. So those are the two premier products that we're using to connect the hardware.

    But also, to generate that 6G exploration signal, MathWorks has a 6G Exploration Library that if you have 5G Toolbox, you can download this library for free. So now let's talk a little bit about these graphics. In the upper left, you see a transmit spectrum of an 800 megahertz OFDM signal. So you see it's nice and flat and beautiful, nice, steep skirts off the stop band. On the right side, that is the received spectrum.

    So you see a little bit of a DC null there at 0. You see a little spur over on the left and then some very noticeable roll off on the right side. All that says is that even though we're going over a wire, that is so broadband that you're still going to see some distortion.

    So an improvement opportunity is to use a really broad beam antenna and see if we can get as flat a passband as possible on the receive end as we have on the transmit end. And then, finally, the constellation diagram on the lower left, that is the received constellation diagram with the impairments that you see spectrally on the right. So you can certainly identify this as a 16-point signal, but it's not pinpoints. You've got serious clouds there. So again, opportunities for making this even better.

    And then if you look at the lower right-hand side on the command window with the scrolling, you can see that the RMS EVM is roughly 13 or so percent, which is not bad. But we look forward to hacking away more at this example as demo and making it even better.

    To learn more about the tools used to create this demo, please see these links. Thanks very much. Bye.

    View more related videos