• DocumentCode
    449525
  • Title

    Signal design for trellis-coded unitary space-time modulation

  • Author

    Wu, Yi ; Lau, Vincent ; Pätzold, Matthias

  • Author_Institution
    Dept. of Inf. & Commun. Technol., Agder Univ. Coll., Norway
  • Volume
    3
  • fYear
    2005
  • fDate
    28 Nov.-2 Dec. 2005
  • Abstract
    In this paper, we present a new formulation of the problem of signal design for trellis-coded unitary space-time communication systems. Our approach differs from other methods in that we not only combine the constellation design and set partitioning process together, we also use directly the Chernoff bound of the pairwise probability of error as design metric, which is a better figure of merit than the previously used chordal distance, for unitary constellation design. By novelly employing a theorem regarding the Clarke subdifferential of the sum of the k largest singular values of the unitary matrix, we present a numerical optimization procedure for finding good constellations and report the best signal constellations found. Simulation results show that these constellations achieve 1 dB gain against the previously used constellations of unitary space-time signals.
  • Keywords
    error statistics; matrix algebra; optimisation; space-time codes; trellis coded modulation; 1 dB; Chernoff bound; Clarke subdifferential; numerical optimization procedure; pairwise error probability; set partitioning process; signal constellations; signal design; trellis-coded unitary space-time modulation; unitary constellation design; unitary matrix; Algorithm design and analysis; Bit error rate; Communication systems; Communications technology; Constellation diagram; Covariance matrix; Educational institutions; Receiving antennas; Signal design; Transmitting antennas;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Global Telecommunications Conference, 2005. GLOBECOM '05. IEEE
  • Print_ISBN
    0-7803-9414-3
  • Type

    conf

  • DOI
    10.1109/GLOCOM.2005.1577916
  • Filename
    1577916