• DocumentCode
    1735980
  • Title

    An efficient MIMO detector algorithm for next generation wireless communication

  • Author

    Xie, Limei

  • Author_Institution
    Sch. of Manage., Changchun Inst. of Technol., Changchun, China
  • Volume
    2
  • fYear
    2011
  • Firstpage
    1195
  • Lastpage
    1199
  • Abstract
    Design a high throughput MIMO-OFDM transceiver and the linear MMSE MIMO detector on a resource limited hardware platform for mobile device is very challenging. In this paper, we study the linear MMSE detector problem and propose a realistic algorithm implementation for a linear MMSE MIMO detector on real hardware. We observed the scaling invariant property of QR decomposition and proposed a dynamic scaling algorithm for QR decomposition. Our algorithm maintains the data in the same range during the factorization process. Our algorithm is very amicable for a fixed point implementation on a hardware platform without floating point units (i.e. low-end FPGA). We have successfully implemented and synthesized the algorithm on Xilinx Vertex-4 FPGAs and shows good numerical stability. Our desgin is very general and applicable to other future MMSE-based MIMO detector algorithm (i.e. 8 × 8 MIMO).
  • Keywords
    MIMO communication; field programmable gate arrays; least mean squares methods; matrix decomposition; mobile handsets; next generation networks; numerical stability; radio transceivers; QR decomposition; Xilinx Vertex-4 FPGA; dynamic scaling algorithm; factorization process; fixed point implementation; floating point units; high throughput MIMO-OFDM transceiver; linear MMSE MIMO detector algorithm; mobile device; next generation wireless communication; numerical stability; realistic algorithm implementation; resource limited hardware platform; scaling invariant property; Algorithm design and analysis; Decoding; Detectors; Heuristic algorithms; MIMO; Matrix decomposition; Vectors; Channel Estimation; FPGA; MIMO; RF-Communication system;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Computer Science and Network Technology (ICCSNT), 2011 International Conference on
  • Conference_Location
    Harbin
  • Print_ISBN
    978-1-4577-1586-0
  • Type

    conf

  • DOI
    10.1109/ICCSNT.2011.6182173
  • Filename
    6182173