Causality-constrained interpolation of tabulated frequency responses

We apply a recently developed formulation of the generalized Hilbert transform to the processing of tabulated and finite-bandwidth frequency responses. A causality-constrained interpolation procedure is introduced, with the aim of reconstructing missing samples in the data, like e.g. the DC point, via a sound numerical procedure that does not compromise the self-consistency and the causality of the entire dataset