• DocumentCode
    2061718
  • Title

    Accumulate repeat accumulate codes

  • Author

    Abbasfar, Aliazam ; Divsalar, Dariush ; Yao, Kung

  • Author_Institution
    Dept. of Electr. Eng., California Univ., Los Angeles, CA, USA
  • fYear
    2004
  • fDate
    2004
  • Firstpage
    505
  • Abstract
    An innovative channel coding scheme called "accumulate repeat accumulate codes" (ARA) is proposed. ARA codes can be viewed as a subclass of low density parity check (LDPC) codes with fast encoder, and they have a projected graph or protograph representation. Using density evolution on their associated protographs, we find examples of rate 1/2 ARA codes with maximum variable node degree 5 for which a minimum bit SNR as low as 0.08 dB from channel capacity can be achieved as the block size goes to infinity. A family of high rate ARA codes with thresholds that stay uniformly close to their respective channel capacity thresholds are constructed. The ensemble weight distribution and ML threshold for rate 1/2 ARA codes were computed. For ARA with repeat 4, the ML threshold approaches within 0.005 dB of the ML threshold of random codes based on the existing tightest closed form bound.
  • Keywords
    channel capacity; channel coding; graph theory; parity check codes; random codes; ARA; LDPC; ML threshold; accumulate repeat accumulate code; bit SNR; channel capacity; density evolution; encoder; ensemble weight distribution; innovative channel coding; low density parity check code; maximum variable node degree; projected graph; protograph representation; random code; Channel capacity; Channel coding; Design methodology; Distributed computing; Gain; H infinity control; Iterative decoding; Laboratories; Parity check codes; Propulsion;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Information Theory, 2004. ISIT 2004. Proceedings. International Symposium on
  • Conference_Location
    Chicago, IL, USA
  • Print_ISBN
    0-7803-8280-3
  • Type

    conf

  • DOI
    10.1109/ISIT.2004.1365542
  • Filename
    1365542