Sinusoidal Fitting Decomposition for Instantaneous Characteristic Representation of Multi-Componential Signal

Area of Science:

• Signal Processing
• Non-stationary Signal Analysis
• Time-Frequency Analysis

Background:

• Extracting instantaneous characteristics from non-stationary signals is a significant challenge.
• Existing methods often rely on dictionaries or basis functions, limiting their applicability.
• Accurate decomposition is crucial for understanding complex signal behaviors.

Purpose of the Study:

• To propose a new multi-component decomposition method called Sinusoidal Fitting Decomposition (SFD).
• To extract Instantaneous Amplitude (IA), Instantaneous Phase (IP), and Instantaneous Frequency (IF) from signals.
• To ensure extracted components possess desirable properties like positive IA and monotonic IP.

Main Methods:

• Developed Sinusoidal Fitting Decomposition (SFD), a novel iterative synthesis process.
• SFD decomposes signals into finite mono-component signals.
• The method avoids reliance on dictionaries, basis function spaces, or sifting operations.

Main Results:

• SFD ensures extracted IA is positive and IP is monotonically increasing.
• The synthesized signal from IA and IP is mono-componential and smooth.
• The method effectively describes instantaneous frequency-amplitude characteristics on the time-frequency plane.

Conclusions:

• Sinusoidal Fitting Decomposition (SFD) is an effective method for non-stationary signal analysis.
• The proposed method offers advantages over existing techniques by avoiding complex dependencies.
• SFD provides accurate and reliable extraction of signal's instantaneous attributes.