• DocumentCode
    2504169
  • Title

    Blockwise Uniform Channel Decomposition for MIMO Systems

  • Author

    Lee, Kyoung-Jae ; Hochwald, Bertrand M. ; Lee, Inkyu

  • Author_Institution
    Korea Univ., Seoul
  • fYear
    2007
  • fDate
    26-30 Nov. 2007
  • Firstpage
    3154
  • Lastpage
    3158
  • Abstract
    In this paper, we investigate spatial multiplexing schemes for closed-loop multiple-input multiple-output (MIMO) systems. The performance of the singular value decomposition (SVD) scheme is limited by the smallest singular value. When all the subchannels are utilized, uniform channel decomposition (UCD) was recently proposed to obtain a performance gain by making subchannels have equal gains. The UCD requires a successive interference cancellation (SIC) receiver, and thus it suffers from the error propagation inherent in the SIC receiver. We propose the blockwise UCD (BL-UCD) scheme which increases the minimum subchannel gain by pairing two singular values. The proposed scheme allows single-symbol decodable maximum-likelihood detection (MLD) instead of the SIC receiver. The simulation results demonstrate that the proposed BL-UCD scheme outperforms the SVD scheme and the conventional UCD at full spatial multiplexing for four transmit antennas and four receive antennas by 8 dB and 5 dB, respectively.
  • Keywords
    MIMO communication; channel coding; diversity reception; interference suppression; maximum likelihood detection; radio receivers; MIMO systems; blockwise uniform channel decomposition; error propagation; maximum-likelihood detection; performance gain; singular value decomposition; spatial multiplexing schemes; Interference cancellation; MIMO; Maximum likelihood decoding; Maximum likelihood detection; Performance gain; Receiving antennas; Silicon carbide; Singular value decomposition; Transmitting antennas; User centered design;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Global Telecommunications Conference, 2007. GLOBECOM '07. IEEE
  • Conference_Location
    Washington, DC
  • Print_ISBN
    978-1-4244-1042-2
  • Electronic_ISBN
    978-1-4244-1043-9
  • Type

    conf

  • DOI
    10.1109/GLOCOM.2007.597
  • Filename
    4411507