Adaptive Constraints-Based Instantaneous Frequency Estimation for Overlapped Multicomponent Signals

Accurate extraction of instantaneous frequency (IF) from overlapped non-stationary signals is essential for applications in radar, communications, and biomedical engineering. Traditional methods like Ridge Path Regrouping (RPRG) introduce artificial artifacts through post-processing, limiting their effectiveness in complex signal environments. This study presents the Adaptive Ridge Path Regrouping (ARPRG) algorithm, which overcomes these limitations by using a generalized regulation function as a time-frequency (TF) filter to constrain IF variation to provide a first order approximation for the IFs with adaptive approximation horizon. A genetic algorithm is used to adaptively optimize the approximation horizon, providing coarse IF estimates based on energy entropy without expert intervention. A newly developed penalty function, derived from linear least squares fitting, further refines the IF in overlapping TF regions. Simulations and microDoppler experiments show that ARPRG significantly improves IF extraction accuracy in overlapped signals, outperforming traditional methods. This algorithm offers a robust solution for complex signal environments, with broad applications in advanced signal processing.