Multiharmonic tracking using sigma-point Kalman filter.

Area of Science:

• Signal Processing
• Time-Frequency Analysis
• State-Space Methods

Background:

• Multi-harmonic quasi-periodic signals are common in various scientific fields.
• Accurate tracking of time-varying frequencies in these signals is crucial for analysis.
• White Gaussian noise often contaminates real-world signals, posing a challenge for frequency estimation.

Purpose of the Study:

• To compare the performance of the extended Kalman filter (EKF) and sigma-point Kalman filter (SPKF) for tracking time-varying frequencies.
• To evaluate the effectiveness of these algorithms under varying signal-to-noise ratios (SNR).

Main Methods:

• State-space modeling was employed to represent the signal dynamics.
• The extended Kalman filter (EKF) algorithm was implemented for frequency tracking.
• The sigma-point Kalman filter (SPKF) algorithm was implemented for frequency tracking.
• Performance was assessed by comparing the accuracy of instantaneous frequency estimation.

Main Results:

• The sigma-point Kalman filter (SPKF) demonstrated superior performance compared to the extended Kalman filter (EKF).
• SPKF achieved more accurate tracking of instantaneous frequencies across a broad spectrum of signal-to-noise ratios (SNR).

Conclusions:

• The sigma-point Kalman filter (SPKF) is a more effective tool for tracking time-varying frequencies in noisy multi-harmonic quasi-periodic signals.
• SPKF offers enhanced accuracy and robustness, making it suitable for applications with challenging signal conditions.