An efficient algorithm for Gaussian-based signal decomposition

Parameter estimation of each atom in the adaptive Gaussian basis representation (AGR) is a key problem determining signal decomposition results, which requires an effective parameter estimation algorithm. An efficient algorithm is proposed to estimate the time-frequency atoms in AGR. Compared with the algorithm presented in AGR, the parameter estimation is more adaptive, and the estimation accuracy is greatly improved through an iterative method. Numerical simulations confirm the results.