Autodetect_parameter suggestion for modulation detection

[lucas12675]

!! I am not an english native writter

Error in the prediction because the signal is recorded with high sample rate

Expected Behavior

The auto-detection is guessing wrong prediction when using high sample rate. The modulation of an FSK signal is wrongfully guessed as OOK if recorded with a sampling_rate of 8MHz signal but the guess is right for the same signal resampled at 1 Mhz

Actual Behavior

The modulation of an FSK signal is wrongfully guessed as OOK.

Steps To Reproduce

1. Record a FSK signal at 8MHz (for me 50 microseconds bits)
2. import it
3. bandpass around it
4. use the autodetect button

Fix suggestion

After investigation the wavelet criteria work using the variance but a signal over-sampled is "less variant". So the first condition in AutoInterpretation.py | detect_modulation, should be adaptative to the sampling_rate :
if all(
v < 0.15 / (sample_rate/1000000)**2
for v in (var_mag, var_norm_mag, var_filtered_mag, var_filtered_norm_mag)
):
(Using the sampling_rate if given in the Signal class)

The square is from the formula of the variance calculus of random variable multiplied by a scalar

Sorry if my english is lacking, I hope my suggestion is constructive for the project