• DocumentCode
    1312113
  • Title

    The Universality of Generalized Hamming Code for Multiple Sources

  • Author

    Ma, Rick ; Cheng, Samuel

  • Author_Institution
    Dept. of Math., Hong Kong Univ. of Sci. & Technol., Hong Kong, China
  • Volume
    59
  • Issue
    10
  • fYear
    2011
  • fDate
    10/1/2011 12:00:00 AM
  • Firstpage
    2641
  • Lastpage
    2647
  • Abstract
    We consider zero-error Slepian-Wolf coding for a special kind of correlated sources known as Hamming sources. Moreover, we focus on the design of codes with minimum redundancy (i.e., perfect codes). As shown in a prior work by Koulgi et al., the design of a perfect code for a general source is very difficult and in fact is NP-hard. In our recent work, we introduce a subset of perfect codes for Hamming sources known as Hamming Codes for Multiple Sources (HCMSs). In this work, we extend HCMSs to generalized HCMSs, which can be proved to include all perfect codes for Hamming sources. To prove our main result, we first show that any perfect code for a Hamming source with two terminals is equivalent to a Hamming code for asymmetric Slepian Wolf coding (c.f. Lemma 2). We then show that any multi-terminal (of more than two terminals) perfect code can be transformed to a perfect code for two terminals (c.f. Lemma 3) and to a perfect code with an asymmetric form (c.f. Lemma 4). Equipped with these results, we prove that every perfect Slepian-Wolf code for Hamming sources is equivalent to a generalized HCMS.
  • Keywords
    Hamming codes; optimisation; source coding; HCMS; Hamming codes for multiple sources; NP-hard problem; asymmetric Slepian Wolf coding; generalized Hamming code; zero-error source coding; Base stations; Channel coding; Decoding; Entropy; Matrix decomposition; Redundancy; Hamming code; Slepian-Wolf coding; zero-error source coding;
  • fLanguage
    English
  • Journal_Title
    Communications, IEEE Transactions on
  • Publisher
    ieee
  • ISSN
    0090-6778
  • Type

    jour

  • DOI
    10.1109/TCOMM.2011.081711.100211
  • Filename
    6007024