Spectral Signal Denoising Algorithm Based On Modified Singular Spectrum Analysis Analysis

A new singular spectrum analysis denoising algorithm is proposed to effectively remove photon noise and detector noise in fiber Bragg gratings. This algorithm represents a one-dimensional spectral signal as a trajectory matrix and obtains eigenvalues and eigenvectors through singular value decomposition. In the eigenvalues vector, singular values are selected through a modified grouping scheme, and finally, the processed signal matrix is reconstructed using the eigenvectors of the selected singular values. The simulation results show that the proposed algorithm can effectively remove noise and improve the signal-to-noise ratio by 8-10dB compared with traditional methods.