Estimating the Probability Density Function of a Nonstationary Non-Gaussian Noise

The problem of estimating the probability density function (pdf) of a nonstationary non-Gaussian noise is addressed. The non-Gaussian noise is modeled using Gaussian mixture pdfs, and an algorithm is proposed to estimate the parameters by maximizing the log-likelihood function. Three simulation results illustrate the validity and utility of the proposed algorithm for stationary or nonstationary, Gaussian or non-Gaussian, zero mean or nonzero mean, and unimodal or multimodal distributed noise.