IF Estimation in Multi-Component Signals Via Hough-Viterbi Algorithm With HGA-ADTFD

Instantaneous frequency estimation is significant in signal processing for non-stationary multi-component signals in dense time-frequency environments. Researchers studied different algorithms for instantaneous frequency to suppress frequency switching errors. However, many classical algorithms are susceptible to errors in the time-frequency distribution (TFD) and noise at instantaneous frequencies, which often degrades their performance for multi-component frequency estimation. We propose a hybrid genetic algorithm-optimized adaptive directional time-frequency distribution (HGA-ADTFD) to improve the efficiency of multi-signal processing in TFD. The proposed HGA-ADTFD algorithm optimizes TFD using genetic parameter optimization. Crucially, the algorithm implements a joint calibration strategy that fuses local linear segments from the Hough transform with continuous paths from the Viterbi algorithm to determine the final instantaneous frequency. The proposed HGA-ADTFD algorithm achieves a minimum root mean square error for the final instantaneous frequency value, which is reduced to 0.5 MHz (the sampling frequency is 100 MHz).