• DocumentCode
    2426078
  • Title

    Expanded Soft Demapper for LDPC Coded GMD-THP MIMO System

  • Author

    Peh, Edward C Y ; Liang, Ying-Chang

  • Author_Institution
    Inst. for Infocomm Res.
  • fYear
    2007
  • fDate
    9-11 Jan. 2007
  • Firstpage
    519
  • Lastpage
    522
  • Abstract
    Tomlinson-Harashima precoding (THP) can be combined with geometric mean decomposition (GMD) to decouple a multiple input multiple output (MIMO) channel into multiple subchannels with identical signal-to-noise ratio (SNR). The combined system is called GMD-THP MIMO system. As the subchannels have identical SNR, it is more convenient for GMD-THP MIMO systems to design modulation/demodulation and coding/decoding schemes than singular value decomposition (SVD) MIMO systems. In this paper, we consider low-density parity-check (LDPC) coded GMD-THP MIMO system. Since the modulo operation at the receiver may generate modulo errors, the log likelihood ratio (LLR) values provided by conventional soft demapper can be very inaccurate. To overcome this problem, we propose an expanded soft demapper scheme to obtain better LLR values, by which significant performance improvement can be achieved for LDPC decoding as compared to conventional soft demapper. Computer simulations are presented to show that the expanded soft demapper indeed improves the error rate performance of the LDPC coded GMD-THP systems and reduces the noise floor caused by the modulo errors
  • Keywords
    MIMO communication; channel coding; decoding; demodulation; parity check codes; precoding; singular value decomposition; wireless channels; GMD-THP MIMO; LDPC decoding; SNR; Tomlinson-Harashima precoding; decoding schemes; demodulation; expanded soft demapper; geometric mean decomposition; log likelihood ratio; low-density parity-check; multiple input multiple output channel; signal-to-noise ratio; singular value decomposition; Computer errors; Computer simulation; Decoding; Demodulation; Error analysis; MIMO; Modulation coding; Parity check codes; Signal to noise ratio; Singular value decomposition;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Radio and Wireless Symposium, 2007 IEEE
  • Conference_Location
    Long Beach, CA
  • Print_ISBN
    1-4244-0445-2
  • Electronic_ISBN
    1-4244-0445-2
  • Type

    conf

  • DOI
    10.1109/RWS.2007.351882
  • Filename
    4160765