• DocumentCode
    1858894
  • Title

    Making Attractive Channel Estimation Method Using Truncated SVD for MIMO-OFDM Systems Thanks to Large Complexity Reduction

  • Author

    Diallo, Moussa ; Rabineau, Rodrigue ; Cariou, Laurent ; Hélard, Maryline

  • Author_Institution
    R&D Div., France Telecom, Cesson-Sevigne, France
  • fYear
    2010
  • fDate
    2-5 Aug. 2010
  • Firstpage
    1
  • Lastpage
    5
  • Abstract
    In this paper we consider channel estimation for MIMO-OFDM systems, specifically when null carriers are placed at the border of the spectrum in order to limit interference with neighboring spectrums. Indeed, the presence of these null carriers leads to degradation on the transform domain channel estimation (TD-CE) because of a "border effect" phenomenon. Truncated singular value decomposition (TSVD) based method allows to improve the accuracy of the TD-CE by significantly reducing each "border effect". However, its application to MIMO systems may cause problem due to its high complexity. We propose in this paper, a solution based on the power profile of the transfer matrix in order to considerably reduce the complexity. The efficiency of the proposed solution is demonstrated within 3GPP system context.
  • Keywords
    3G mobile communication; MIMO communication; OFDM modulation; channel estimation; computational complexity; singular value decomposition; 3GPP system context; MIMO-OFDM systems; complexity reduction; transfer matrix; transform domain channel estimation; truncated SVD method; truncated singular value decomposition; Channel estimation; Complexity theory; Discrete Fourier transforms; Discrete cosine transforms; Frequency domain analysis; MIMO; OFDM;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Computer Communications and Networks (ICCCN), 2010 Proceedings of 19th International Conference on
  • Conference_Location
    Zurich
  • ISSN
    1095-2055
  • Print_ISBN
    978-1-4244-7114-0
  • Type

    conf

  • DOI
    10.1109/ICCCN.2010.5560148
  • Filename
    5560148