• DocumentCode
    2429505
  • Title

    Vector precoding based on Geometric Mean Decomposition for MIMO transmission system

  • Author

    Geng, Xuan ; Jiang, Lingge ; He, Chen

  • Author_Institution
    Dept. of Electron. Eng., Shanghai Jiao Tong Univ., Shanghai
  • fYear
    2008
  • fDate
    7-11 June 2008
  • Firstpage
    280
  • Lastpage
    284
  • Abstract
    The Geometric Mean Decomposition (GMD) for MIMO channel matrix can obtain identical subbchannel gains, which is a useful property to improve performance gains of precoding. In this paper, we combine the vector precoding with GMD for channel matrix to design the transceiver for MIMO transmission system. Under the proposed transceiver structure, the minimum square error (MSE) of receive symbols are obtained. To minimize the MSE, two schemes in terms of perturbation vector are proposed. In the first scheme, the perturbation vector has only continuous values, thus it is regarded as interference for MSE. In the second scheme, the perturbation vector is generalized to have both continuous and discrete values, so it is partially dealt with as interference after modulo operation. In these two schemes, the optimum perturbation vectors are presented in MMSE criterion respectively.
  • Keywords
    MIMO communication; least mean squares methods; matrix algebra; precoding; transceivers; vectors; wireless channels; MIMO transmission system; MMSE; channel matrix; geometric mean decomposition; minimum square error method; perturbation vector; transceiver design; vector precoding; MIMO; Matrix decomposition; Neural networks; Performance gain; Performance loss; Radiofrequency interference; Signal processing; Transceivers; Transmitters; Wireless communication; GMD; MIMO; Vector precoding;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Neural Networks and Signal Processing, 2008 International Conference on
  • Conference_Location
    Nanjing
  • Print_ISBN
    978-1-4244-2310-1
  • Electronic_ISBN
    978-1-4244-2311-8
  • Type

    conf

  • DOI
    10.1109/ICNNSP.2008.4590356
  • Filename
    4590356