Range FFT Magnitude Too High – Help Needed to Get Proper Y-Axis Magnitude

Hi everyone,

I'm encountering an issue with my radar signal processing code in MATLAB. Specifically, the magnitude values after applying the Range FFT are much higher than expected, even after windowing and normalization.

Could someone please help me understand what might be causing this? I'm trying to get the correct power values in dBm on the y-axis.

I’ve attached the relevant MATLAB code and data files for reference. The issue can be seen clearly in the output plot where the y-axis magnitude appears significantly higher than expected, despite the input power being in negative dBm range.

If anyone can suggest what steps or corrections are needed to achieve the proper magnitude representation after FFT, I would greatly appreciate it. Thanks in advance!

[Star Strider]

I can't find any reference to a 'range_fft' function in the online documentation, so I assume it's something you wrote.

We also don't have 'filtered_signal' so that makes everything else irrelevant anyway.

[A]

Hi @Star Strider, thank you for your response.

Just to clarify, my FMCW radar processing flow is as follows:

Raw data is captured from the ADC.

This ADC data is passed through a FIR low-pass filter with a cutoff frequency of 10MHz (using fir1).

The filtered output is then fed into the Range FFT stage for further processing.

That's the reason the filtered signal appears in the pipeline before FFT. I wanted to ensure this was clear for context. If you have any suggestions on how this filtering might affect the magnitude scaling or interpretation in dBm, I’d really appreciate your insights.

Thanks again for your support!

%% FIR Filtering 
cutoff = 10e6 / (fs/2);
fir_order = 32; 
fir_coeff = fir1(fir_order, cutoff, 'low'); 
filtered_signal = filter(fir_coeff, 1, adc_data.');

[Matt J]

Please run your code for us here in the online environment, so we can only agree on what it shows. Example,

fs = 100e6;
num_samples = 2048;   
window_coeff = hamming(num_samples).';

plot(window_coeff)

Answer 1

[Star Strider]

@A --

With respect to your earlier Comment, a FIR filter is not the best choice if you need only one filter. I would suggest instead using the lowpass function:

filtered_signal = lowpass(adc_data, cutoff, fs, ImpulseResponse='iir');

That will be much more efficient (it designs a very efficient elliptic flter and then uses filtfilt to do the actual filtering) and produces a better result.

We still don't have the 'range_fft' function. For what it's worth, I wrote my own single-sided FFT functiona while back for my own use that you can consider using:

function [FTs1,Fv] = FFT1(s,t)

% One-Sided Numerical Fourier Transform

% Arguments:

% s: Signal Vector Or Matrix

% t: Associated Time Vector

t = t(:);

L = numel(t);

if size(s,2) == L

s = s.';

end

Fs = 1/mean(diff(t));

Fn = Fs/2;

NFFT = 2^nextpow2(L);

FTs = fft((s - mean(s)) .* hann(L).*ones(1,size(s,2)), NFFT)/sum(hann(L));

Fv = Fs*(0:(NFFT/2))/NFFT;

% Fv = linspace(0, 1, NFFT/2+1)*Fn;

Iv = 1:numel(Fv);

Fv = Fv(:);

FTs1 = FTs(Iv,:);

end

Note that 's' will be converted to a column vector or column-major matrix and then analysed, so the results will be a column vector or column-major matrix.

To plot the resutl:

figure

plot(Fv, abs(FTs1)*2)

grid

xlabel('Frequency (units)')

ylabel('Magnitude (units)')

or:

figure

plot(Fv, mag2db(abs(FTs1)*2))

grid

xlabel('Frequency (units)')

ylabel('Power (dB)')

Experiment with this to see if it produces the result you want. If it does not, please post back with comments and suggestions.

.