Exact equivalence of the Steiglitz-McBride iteration and IQML

Abstract

The exact equivalence of the Steiglitz-McBride mode-1 algorithm and the iterative quadratic maximum likelihood algorithm (IQML) is demonstrated. Other methods, published earlier by L. L. Scharf et al. (1986) and by A.G. Evans and R. Fischl (1973), are also known to give the same equations and answer as IQML. Hence, all these methods are equivalent to the Steiglitz-McBride iteration. Since the IQML method was originally formulated to solve a multiple measurement case, an extended version of the Steiglitz-McBride mode-1 algorithm, applicable to the multiple measurement case, is also stated and shown to be equivalent to the general form of IQML. >