• DocumentCode
    2640713
  • Title

    A Modified LMS Algorithm with Turbo-Like Scheme

  • Author

    Liu, Qian ; Jin, Minglu ; Liu, Aifei

  • Author_Institution
    Dept. of Electron. Eng., Dalian Univ. of Technol., Dalian
  • fYear
    2008
  • fDate
    18-20 June 2008
  • Firstpage
    534
  • Lastpage
    534
  • Abstract
    LMS algorithm has low computation complexity; however, its convergence rate is slow. In this paper, a novel implementation scheme for LMS algorithm is proposed to form a modified LMS algorithm. The scheme is similar to the structure of turbo decoder, which consists of one feedback module called as converse-speediness module, two LMS algorithm modules, and one speediness module which connects two LMS algorithm modules. Meanwhile, two LMS algorithm modules can be implemented parallel, as well as the speediness module and converse-speediness module. Simulation results show that the scheme makes the convergence rate of the proposed algorithm over 4 times faster than that of LMS algorithm, and almost the same as that of RLS algorithm. Moreover, compared to LMS algorithm and RLS algorithm, the mean square error (MSE) of the proposed algorithm decreases greatly.
  • Keywords
    computational complexity; decoding; filtering theory; least mean squares methods; turbo codes; LMS algorithm; MSE; computation complexity; converse-speediness module; feedback module; least mean squares filter; mean square error; turbo decoder; turbo-like schemes; Computational modeling; Convergence; Decoding; Displays; Feedback; Iterative algorithms; Least squares approximation; Mean square error methods; Resonance light scattering; Samarium;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Innovative Computing Information and Control, 2008. ICICIC '08. 3rd International Conference on
  • Conference_Location
    Dalian, Liaoning
  • Print_ISBN
    978-0-7695-3161-8
  • Electronic_ISBN
    978-0-7695-3161-8
  • Type

    conf

  • DOI
    10.1109/ICICIC.2008.43
  • Filename
    4603723