Enhanced Adaptive Chirp Mode Decomposition With Instantaneous Frequency Refinement: A Micro-Doppler Processing Method

Passive radar systems have gained significant attention due to their stealth operation and infrastructure-free implementation. Such systems typically rely on micro-Doppler (MD) features generated from micromotions of structural components such as rotating blades as key discriminative features for target recognition. However, the time–frequency representation (TFR) becomes blurry due to noise contamination and complex superposition of instantaneous frequency (IF) components, presenting significant challenges for passive radar systems in analyzing MD signals (MDSs). To address these limitations, this article proposes an enhanced adaptive chirp mode decomposition (ACMD) method incorporating sparsity-based IF optimization. By introducing IF smoothing terms and sparse regularization of their spectrum in the ACMD framework, it is possible to suppress abrupt changes caused by noise and modes far from the IF, as well as oscillations caused by noise around the IF. Both synthetic and real-world data validations demonstrate the effectiveness of the proposed method in extracting MDSs from multiple IF components’ superposition scenarios.