Matched power-frequency-modulated signal transform and its application in bat call signal analysis

Area of Science:

• Bioacoustics
• Signal Processing
• Animal Communication

Background:

• Bat call analysis is crucial for species identification and biomimetic systems.
• Traditional time-frequency methods and the fractional Fourier transform (FRFT) have limitations in analyzing nonlinear bat call modulations.

Purpose of the Study:

• To propose a new integral transform, the Matched Power-Frequency-Modulated (PFM) signal transform (MPST), for analyzing nonlinear bat calls.
• To address the limitations of FRFT in processing bat calls with nonlinear frequency modulation.

Main Methods:

• The study introduces the MPST, a time-warping Fourier transform generalization.
• MPST is designed for analyzing power-frequency-modulated (PFM) signals, where frequency is an approximate power function of time.
• The transform's performance is validated on European bat call recordings.

Main Results:

• MPST effectively analyzes bat calls modeled with PFM characteristics.
• The transform facilitates parameter estimation and harmonic separation in bat vocalizations.
• Validated performance across feeding buzzes, social calls, and distress calls.

Conclusions:

• The proposed MPST is a valuable tool for analyzing complex, nonlinear bat vocalizations.
• MPST offers improved capabilities over FRFT for specific bat call structures.
• This advancement supports more accurate bat species identification and biomimetic applications.