A multistage representation of the Wiener filter based on orthogonal projections

Abstract

The Wiener filter is analyzed for stationary complex Gaussian signals from an information-theoretic point of view. A dual-port analysis of the Wiener filter leads to a decomposition based on orthogonal projections and results in a new multistage method for implementing the Wiener filter using a nested chain of scalar Wiener filters. This new representation of the Wiener filter provides the capability to perform an information-theoretic analysis of previous, basis-dependent, reduced-rank Wiener filters. This analysis demonstrates that the recently introduced cross-spectral metric is optimal in the sense that it maximizes mutual information between the observed and desired processes. A new reduced-rank Wiener filter is developed based on this new structure which evolves a basis using successive projections of the desired signal onto orthogonal, lower dimensional subspaces. The performance is evaluated using a comparative computer analysis model and it is demonstrated that the low-complexity multistage reduced-rank Wiener filter is capable of outperforming the more complex eigendecomposition-based methods.

Performance of Interleave Division Multiple Access Based on Minimum Mean Square Error Detection

“…A multistage representation of the MMSE or Wiener filter (WF) based on orthogonal projections was introduced in [14]. The solution is called multistage WF (MSWF).…”

A robust BFCC feature extraction for ASR system

“…Although MFCC was investigated the sophisticated performance in speech recognition system, the properties of Fourier transform and the triangular mel-filterbank in MFCC were shown unlikely the sound wave sensitivity at basilar-membrane in human auditory system, and gave the less robustness in the presence of additive noise. In other works, suppression of the additive noise interruption for ASR by using spectral subtraction methods [5] and by Wiener filters [6] were ever well-investigated.…”

Exploiting persymmetry for low-rank Space Time Adaptive Processing

“…To avoid the compu-1 LR-STAP filters built from other estimators (like for instance Normalized SCM) are less per-2 tation of the SVD and to limit the computational time, some algorithms [15,16] based on subspace tracking has been proposed. Similarly, to fill these gaps, new STAP algorithms based on a projection received data onto a lower dimensional Krylov subspace [17,18,19,20] or based on joint iterative optimization of adaptive filters [21,22] have been recently developed 2 . In this paper, we derive a new STAP filter from the SVD of a new estimator of the CM in order to still reduce the number of secondary data by reaching the same performance.…”