# How to model Additional EVM (TxEVM) with LTE waveform?

I would like mention my understanding in the following,

In an LTE OFDM signal assume, Nfft=8; scs=15 kHz; BW=5MHz; Sampling Rate N=7.68 MHz; We are taking 1msec of sample; Therefore total sample =76800 sample.

I found a model from Mathworks which is explained on this link.

    txWaveform={76800} % Assume OFDM Signal in time doamin
txEVMpc = 1.2;            % Set desired transmit EVM in percent
gain = txEVMpc/(100*sqrt(double(info.Nfft))); % Compute gain of additive noise
evmModel = gain * complex(randn(size(txWaveform)), randn(size(txWaveform)))/2;
rxWaveform = txWaveform+evmModel;        % Add noise to the transmitted waveform


Now I have the following question,

1. In the code on line 3 the gain it depends on Nfft size (it is 8 in my example here). However, I can have some previous scaling on the "txWavefrom" (I can increase the signal amplitude). In that case 'evmmodel' signal power remains constant but this should be adjusted!!! how?

2. If I do oversampling of txWaveform, how can I adjust the EVM gain?

3. Any link or tutorial on additional EVM modeling would help a lot.

Thank you.