• DocumentCode
    1005417
  • Title

    Multistep prediction of autoregressive signals

  • Author

    Stoica, Petre

  • Author_Institution
    Uppsala Univ., Sweden
  • Volume
    29
  • Issue
    6
  • fYear
    1993
  • fDate
    3/18/1993 12:00:00 AM
  • Firstpage
    554
  • Lastpage
    555
  • Abstract
    The multistep predictions of an autoregressive (AR) signal satisfy a simple recursive equation. The only derivation of this useful property of which the author is aware makes use of the conditional mean methodology. A derivation based on the polynomial approach is presented by the author. The latter approach to the prediction of signals with rational spectra has a number of advantages over the conditional mean approach, which is also briefly discussed.
  • Keywords
    filtering and prediction theory; polynomials; signal processing; autoregressive signals; conditional mean methodology; multistep predictions; rational spectra; recursive equation;
  • fLanguage
    English
  • Journal_Title
    Electronics Letters
  • Publisher
    iet
  • ISSN
    0013-5194
  • Type

    jour

  • DOI
    10.1049/el:19930369
  • Filename
    256238