Pole-zero estimation of continuous and discrete time systems

An approach to pole-zero estimation for mathematical modeling of a system given a finite duration of its response is presented. The basic tool is in the form of weighted Laplace and z transform spectra which are shown to reveal pole peaks normally absent due to exponential rise in spectra on the complex plane. The approach is an extension of a parallel algorithm recently proposed and effectively performs an extrapolation yielding the infinite impulse response of systems given their finite impulse response. Applications of the approach include mathematical model reduction, system identification, adaptive filtering and adaptive equalization among others.