On the use of the minimum phase algorithm in EMC data processing

The phase retrieval problem consisting of estimating the phase of a transfer function from measurement of its magnitude has been around for several years in the EMC community. In the first part of the paper instead of the direct application of the Hilbert integral, two alternative approaches are briefly illustrated: the cepstrum algorithm and the iterative method. They are used in the signal processing algorithm combining the estimation and detection procedures in sequence to identify the structure of an unknown transfer function: in the first step we determine the poles and zeros of the minimum phase transfer function (estimation), in the second step we find out the actual transfer function by comparing its output related to the unknown zero pole pattern with the outputs related to all the possible zero and pole patterns (detection). The structure of the model is particularly suitable to implement the parametric model, such as the Steiglitz Mc Bride algorithm.