Identification of dynamic systems from noisy data; single factor case

Lineardvnamicerrors-in-~ariables~~rfaUor~moddsintheframework of stationary proceareo arc wnsidercd The noise pronss is a~rumed to havc a diagonal srwl~al density. Weanaka the relation betwcen thc (population) socund momentsofthe observakonsandthesystemandnoisecharact&ti(S; ofp.&& interest are the number of equations (or the number of factors) and a dwription of the set of all systems mmGtible with thssecandmo-ts if the observations In this paper emphasis is put on the case which oo be reduced to a single factor. Tbe problems mnsidered arise in the context of identification and noise modeling.