Estimation of Spectral Lines Using Expectation Propagation

We consider the line spectral estimation (LSE) from general linear/nonlinear measurements obtained through a generalized linear model (GLM). This paper develops expectation propagation (EP) based LSE (EPLSE) method. The proposed method automatically estimates the model order, noise variance, and can deal with the nonlinear measurements. Numerical experiments show the excellent performance of EPLSE.