An important problem of digital signal processing is the so-called
frequency analysis problem: Let $f$ be an anharmonic Fourier sum.
Determine the different frequencies, the coefficients, and the
number of frequencies from finitely many equispaced sampled data
of $f$. This is a nonlinear inverse problem. In this paper, we
present new results on an approximate Prony method which is based
on \cite{BeMo02, BeMo05}. In contrast to \cite{BeMo02, BeMo05}, we
apply matrix perturbation theory such that we can describe the
properties and the numerical behavior of the approximate Prony
method in detail. Numerical experiments show the performance of
our method.
Keywords:
frequency analysis
problem, nonequispaced fast Fourier transform, digital signal
processing, anharmonic Fourier sum, approximate Prony method,
matrix perturbation theory, perturbed Hankel matrix,
Vandermonde-type matrix