Frequency domain maximum likelihood estimation of linear dynamic errors-in-variables models

This paper studies the linear dynamic errors-in-variables problem in the frequency domain. First the identifiability is shown under relaxed conditions. Next a frequency domain Gaussian maximum likelihood (ML) estimator is constructed that can handle discrete-time as well as continuous-time models on (a) part(s) of the unit circle or imaginary axis. The ML estimates are calculated via a computational simple and numerical stable Newton-Gauss minimization scheme. Finally the Cramr-Rao lower bound is derived.