Automated frequency band estimation for sinusoidal low-pass signals in noise

Classic signal analysis usually begins with collecting information about the properties of the signal to the greatest extent possible. This information is the basis for the sensible selection of analysis methods and their parameterization.

For sinusoidal signals, the frequency band of the signal is one of these fundamental signal properties. Estimating the frequency band can be a challenge, especially for signals with a small signal-to-noise ratio.

This work proposes a fully automated method for easily, quickly, and stably estimating the frequency band of low-pass signals. The estimation is based on the discrete Fourier transform and threshold detection. Due to the complete automation, special attention is paid to the method's stability.

The performance of the method is tested here using synthetic and natural signals. The preliminary results show that the automated estimation and visual detection deliver comparable results for the upper frequency band limit.