Abstract
Chi and Mendel [1], [9] analyzed the performance of the minimum-variance deconvolution (MVD). In this correspondence, a further analysis of the performance of the MVD filter is presented. We show that the MVD filter performs like an inverse filter and a whitening filter as SNR → ∞, and like a matched filter as SNR -→ 0. The estimation error e (k) of the MVD filter is colored noise, but it becomes white when SNR → 0. This analysis also connects the error power spectral density of the MVD filter with the spectrum of the causal prediction error filter. Copyright © 1987 by The Institute of Electrical and Electronics Engineers, Inc.