• DocumentCode
    1161197
  • Title

    Regularity and unitarity of bilinear time-frequency signal representations

  • Author

    Hlawatsch, Franz

  • Author_Institution
    Inst. fuer Nachrichtentechnik und Hochfrequenztechnik, Tech. Univ. Wien, Austria
  • Volume
    38
  • Issue
    1
  • fYear
    1992
  • fDate
    1/1/1992 12:00:00 AM
  • Firstpage
    82
  • Lastpage
    94
  • Abstract
    Two structural properties of bilinear time-frequency representations (BTFRs) of signals are introduced and studied. The definition of these properties is based on a linear-operator description of BTFRs. The first property, termed regularity, has important implications with respect to the recovery of signals from the BTFR outcome, the derivation of other bilinear signal representations from a BTFR, the BTFRs reaction to linear signal transformations, and the construction of bases of induced BTFR-domain spaces. The second property, called unitarity, is equivalent to validity of Moyal´s formula (1949). Unitarity is thus necessary and sufficient for a closed-form solution of optimal signal synthesis and for a BTFR formulation or optimal detection/estimation methods. Unitarity also allows the systematic construction of BTFR product relations like the Wigner distribution´s interference formula and the ambiguity function´s self-transformation property. Unitarity permits the construction of induced orthogonal projection operators and guarantees the orthonormality of induced basis functions
  • Keywords
    frequency-domain analysis; signal detection; signal processing; signal synthesis; time-domain analysis; Moyal´s formula; Wigner distribution´s interference formula; ambiguity function´s self-transformation property; bilinear signal representations; induced basis functions; induced orthogonal projection operators; linear-operator description; optimal detection/estimation; optimal signal synthesis; regularity; time-frequency signal representations; unitarity; Closed-form solution; Equations; Frequency domain analysis; Interference; Signal analysis; Signal representations; Signal synthesis; Spectrogram; Time domain analysis; Time frequency analysis;
  • fLanguage
    English
  • Journal_Title
    Information Theory, IEEE Transactions on
  • Publisher
    ieee
  • ISSN
    0018-9448
  • Type

    jour

  • DOI
    10.1109/18.108252
  • Filename
    108252