The problem of joint estimation of power spectrum and modulation fromrealizations of frequency modulated stationary wideband signals is considered.The study is motivated by some specific signal classes from which departures tostationarity can carry relevant information and has to be estimated. Theestimation procedure is based upon explicit modeling of the signal as awideband stationary Gaussian signal, transformed by time-dependent, smoothfrequency modulation. Under such assumptions, an approximate expression for thesecond order statistics of the transformed signal's Gabor transform isobtained, which leads to an approximate maximum likelihood estimationprocedure. The proposed approach is validated on numerical simulations.
展开▼