How to use evm?

How do I actually use the comms toolbox evm thing to calculate evm? The Matlab webpages show layers of object-oriented abstraction, but little actual info on how to use it. I just want to use it as a simple function so I don't have to write my own. No need for Simulink. For example, for a QPSK signal, I give it the original constellation and a noisy constellation, and I get back the RMS error. Something like this (greatly simplified):

constellation=[0.7071+0.7071i 0.7071-0.7071i -0.7071+0.7071i -0.7071-0.7071i];
awgn=randn(1,4)+1i*rand(1,4);
err=evm(constellation,constellation+awgn)

Is there any way do that, or any other evm function that does it? Or should I just write my own code?

Thanks.

Answer 1

[Yash]

Hi,

You can use the built-in comm.EVM object as it provides a convenient and standardized way to calculate EVM in MATLAB.

Here is the updated code snippet for the functionality:

constellation = [0.7071+0.7071i, 0.7071-0.7071i, -0.7071+0.7071i, -0.7071-0.7071i]; 
awgn = randn(1, 4) + 1i * rand(1, 4); 

evm = comm.EVM; % creates an EVM measurement system object 
err = evm(constellation, constellation + awgn); 

For more details on using the comm.EVM object and its additional capabilities you can refer to the documentation: Measure error vector magnitude (EVM) of received signal - MATLAB (mathworks.com)

[Jerry Malone]

Thanks a million!

[Jerry Malone]

Is it necessary to recreate the evm object every time I use it?

For example, this works:

    evm=comm.EVM;
    EVMcalc.dist=rms(evm(d,zznMFdss));
    evm=comm.EVM;        
    EVMcalc.cr=rms(evm(d,z_rotated));
    evm=comm.EVM;
    EVMcalc.crrad=rms(evm(exp(1i*angle(z_rotated)),z_rotated));
    evm=comm.EVM;
    EVMcalc.crcirc=rms(evm(d,exp(1i*angle(z_rotated))));
    evm=comm.EVM;
    EVMcalc.eq=rms(evm(d(Npe_start:end),z_eq(Npe_start:end)));
    evm=comm.EVM;
    EVMcalc.dd=rms(evm(d(Npe_start:end),zznMFdss(Npe_start:end)));

But this does not:

    evm=comm.EVM;
    EVMcalc.dist=rms(evm(d,zznMFdss));
    EVMcalc.cr=rms(evm(d,z_rotated));
    EVMcalc.crrad=rms(evm(exp(1i*angle(z_rotated)),z_rotated));
    EVMcalc.crcirc=rms(evm(d,exp(1i*angle(z_rotated))));
    EVMcalc.eq=rms(evm(d(Npe_start:end),z_eq(Npe_start:end)));
    EVMcalc.dd=rms(evm(d(Npe_start:end),zznMFdss(Npe_start:end)));

The last two lines give errors. In all cases I'm giving the evm a reference constellation and a received constellation. I don't see how using it repeatedly changes any of its parameters.