Enhanced Symplectic Geometric Mode Decomposition via K-means Clustering for LFM Signal Denoising

This article presents a Symplectic Geometric Mode Decomposition (SGMD) method incorporating K-means clustering, coupled with wavelet denoising, for mitigating noise in Linear Frequency Modulation (LFM) signals. This method substitutes the original component recombination approach in SGMD with the K-means clustering algorithm, thus rendering the signal grouping more logically coherent. In the experiment, we compare the method proposed in this article with other denoising methods. The experimental findings demonstrate substantial advantages of the proposed method in mitigating noise in LFM signals, thereby enhancing both Signal-to-Noise Ratio (SNR) and signal quality. The research outcomes underscore the feasibility and substantial potential of this method in mitigating noise in signals. In subsequent studies, additional investigation into its applicability across diverse domains can be pursued alongside the refinement of clustering and denoising algorithms.