Asymptotic Uncertainty of Transfer-Function Estimates Using Nonparametric Noise Models

Identification of parametric transfer-function models from noisy input/output observations is an important task in many engineering applications. Aside from the parametric model, the estimation algorithm used should also provide accurate confidence bounds. In addition, it is important to know whether the proposed estimation algorithm has the lowest possible uncertainty within the class of consistent estimators. This paper handles these issues for the frequency-domain Gaussian maximum-likelihood estimator of rational transfer-function models within an errors-invariables framework.