Signal decomposition by the S-method with general window functions

In this letter, we propose a time–frequency signal decomposition algorithm based on the S-method. The proposed algorithm is an extension of the algorithm in [1], so that a general case of the window function used in calculating the S-method can be dealt with. The effect of window function is explicitly verified, based on which a penalized squared error minimization is used to recover the signal matrix. The signal components are reconstructed by eigen decomposition on the recovered signal matrix. Simulation validates the effectiveness of the proposed algorithm.

► Analysis on the effect of general window functions on the S-method.
► Improvement on S-method based signal decomposition with general window functions.
► Multi-component signal processing.