Robust filtering under stochastic parametric uncertainties

Huanshui Zhang, Dapeng Zhang, Lihua Xie, Jun Lin

Research output: Journal article publicationJournal articleAcademic researchpeer-review

43 Citations (Scopus)


This paper is concerned with a polynomial approach to robust deconvolution filtering of linear discrete-time systems with random modeling uncertainties. The modeling errors appear in the coefficients of the numerators and denominators of both the input signal and system transfer function models in the form of random variables with zero means and known upper bounds of the covariances. The robust filtering problem is to find an estimator that minimizes the maximum mean square estimation error over the random parameter uncertainties and input and measurement noises. The key to our solution is to quantify the effect of the random parameter uncertainties by introducing two fictitious noises for which a simple way is given to calculate their covariances. The optimal robust estimator is then computed by solving one spectral factorization and one polynomial equation as in the standard optimal estimator design using a polynomial approach. An example of signal detection in mobile communication is given to illustrate the effectiveness of our approach.
Original languageEnglish
Pages (from-to)1583-1589
Number of pages7
Issue number9
Publication statusPublished - 1 Sept 2004


  • Polynomial approach
  • Robust estimation
  • Signal detection
  • Spectral factorization
  • Stochastic uncertainty

ASJC Scopus subject areas

  • Control and Systems Engineering
  • Electrical and Electronic Engineering


Dive into the research topics of 'Robust filtering under stochastic parametric uncertainties'. Together they form a unique fingerprint.

Cite this