Bimodal histogram of a signal embedded in noise

Hi, Someone could help me about this problem:

I have a signal embedded in White Gaussian Noise (AWGN) with different SNR's. From the histogram it's possible to see that there are two distribution (red and blue), and a want estimate the dispersion around the great one (elipse red) from its great value or its mode.

The figure attached shows a single SNR, but as I change the SNR , the distribution change it's behavior too. Therefore there are aways two peaks well separated.

Someone help please! How I can find this dispersion as an automatic way, since I assume that I don't know nothing about noise, just analising its behavior by means of histogram

Thanks in advance

Answer 1

[Image Analyst]

See the two attached "fit_multiple_gaussians" demo that can fit any number of Gaussians to your data. Adapt the "multiple" demo to use 2 Gaussians instead of 6.

Or use the attached "fit_two_Gaussians" demo.