• DocumentCode
    2171224
  • Title

    Expectation maximization based detection for V-BLAST MIMO communication systems and performance evaluation

  • Author

    Rad, Kamiar Rahnama ; Nasiri-Kenari, Masoumeh

  • Author_Institution
    Sharif Univ. of Technol., Tehran, Iran
  • fYear
    2004
  • fDate
    30 Aug.-2 Sept. 2004
  • Firstpage
    255
  • Lastpage
    259
  • Abstract
    Maximum likelihood naturally arises as the target criterion for detection in many applications, as well as in layered space-time communications. Nevertheless, existing algorithms for maximum likelihood decoding have computational complexity that is not feasible for implementation in practical systems. Applying the expectation maximization algorithm to the maximum likelihood detection of layered space-time codes, we iteratively maximize the conditional log-likelihood of a single layer, rather than maximizing the intractable likelihood function of all layers. Computer simulations demonstrate the efficiency of the proposed detection scheme, in comparison to the well known ing and cancelling detection scheme.
  • Keywords
    MIMO systems; computational complexity; iterative methods; maximum likelihood decoding; maximum likelihood detection; optimisation; radio links; space-time codes; V-BLAST MIMO communication systems; computational complexity; conditional log-likelihood; expectation maximization algorithm; ing-cancelling detection scheme; intractable likelihood function; iterative maximization; layered space-time codes; layered space-time communications; maximum likelihood decoding; maximum likelihood detection; Computational efficiency; Iterative algorithms; Iterative decoding; Laboratories; MIMO; Maximum likelihood decoding; Maximum likelihood detection; Receiving antennas; Space technology; Transmitting antennas;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Spread Spectrum Techniques and Applications, 2004 IEEE Eighth International Symposium on
  • Print_ISBN
    0-7803-8408-3
  • Type

    conf

  • DOI
    10.1109/ISSSTA.2004.1371650
  • Filename
    1371650