• DocumentCode
    1620487
  • Title

    Graph-Matched LDPC Codes for Partial-Response Channels

  • Author

    Legg, Andrei P. ; Uchôa-Filho, Bartolomeu F.

  • Author_Institution
    Dept. of Electr. Eng., Fed. Univ. of Santa Catarina, Florianopolis
  • fYear
    2008
  • Firstpage
    2066
  • Lastpage
    2070
  • Abstract
    In this paper, we propose an LDPC coding scheme based on repeat-accumulate (RA) codes for partial-response (PR) channels. We modify the original RA encoder in such a way that there exists a one-to-one relationship between a sequence called the D-codeword (which is not the transmitted codeword but contains the information bits) and the noiseless received sequence at the output of the PR channel. A consequence of this is that the graph used for decoding the D-codeword, which is different from the encoder graph, is of low complexity. Moreover, with the appropriate modification of the RA encoder, the graph used for decoding is the same of any PR channel, which makes sense calling these codes graph-matched. The simplicity of the decoder, resulting from the "binary" interference removal (precoding) and the linear, as opposed to quadratic, complexity of the encoder make the proposed scheme attractive. For a code rate R = 3641/4096 = 0.89 and for the PR4 channel, an LDPC code has been designed and a simulated coding gain of about 5 dB for a bit error rate less than 10-5 was obtained.
  • Keywords
    Gaussian channels; interference suppression; parity check codes; LDPC coding scheme; binary interference removal; encoder graph; graph-matched LDPC codes; noiseless received sequence; partial-response channels; repeat-accumulate codes; Bit error rate; Communications Society; Convolutional codes; Decoding; Interference; Magnetic recording; Parity check codes; Product codes; Transfer functions; Turbo codes;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Communications, 2008. ICC '08. IEEE International Conference on
  • Conference_Location
    Beijing
  • Print_ISBN
    978-1-4244-2075-9
  • Electronic_ISBN
    978-1-4244-2075-9
  • Type

    conf

  • DOI
    10.1109/ICC.2008.396
  • Filename
    4533432