Spectral-Factorization,-Whitening--and-Estimation-Filter-–-Stability,-Smoothness-Properties-and-FIR-Approximation-Behavior.pdfVIP

Spectral Factorization, Whitening- and Estimation Filter – Stability, Smoothness Properties and FIR Approximation Behavior Holger Boche and Volker Pohl Technical University Berlin, Heinrich Hertz Chair for Mobile Communications Einsteinufer 25, 10587 Berlin, Germany, Email: {holger.boche, volker.pohl}@mk.tu-berlin.de Abstract— A Wiener filter can be interpreted as a cascade corresponding inverse D-transform is defined by of a whitening- and an estimation filter. This paper gives a ∞ detailed investigates of the properties of these two filters. Then k 1 −(k+1) x(z) = xk z and xk = x(z)z dz . (1) the practical consequences for the overall Wiener filter are k=−∞ 2πı |z |=1 ascertained. It is shown that if the given spectral densities are smooth (Hoelder continuous) functions, the resulting Wiener filter Note that the D-transform is the usual Z -transform where z will always be stable and can be approximated arbitrarily well by is replaced by z −1. The D-transform of an impulse response a finite impulse response (FIR) filter. Moreover, the smoothness {hk } is called the transfer function h of the linear system and of the spectral densities characterizes how fast the FIR filter approximates the desired filter characteristic. If on the other its frequency response h(e−ıω ) w