• DocumentCode
    2274471
  • Title

    An unified form of exact-MSR codes via product-matrix framework

  • Author

    Sian-Jheng Lin ; Wei-Ho Chung

  • Author_Institution
    Res. Center for Inf. Technol. Innovation, Acad. Sinica, Taipei, Taiwan
  • fYear
    2013
  • fDate
    8-11 Sept. 2013
  • Firstpage
    830
  • Lastpage
    834
  • Abstract
    Regenerating code represents a class of erasure codes applicable to distributed storage systems. An [n, k, d] regenerating code admits the recovery of the message from any k out of the n encoded fragments, and the recovery of an erasure fragment from other arbitrary d valid fragments, for k ≤ d ≤ n - 1. Minimum Storage Regenerating (MSR) code is the regenerating code attaining the minimal storage requirement. By product-matrix framework, the [n, k, d ≥ 2k-2] exact-MSR code is presented by Rashmi et al. This paper presents a novel version of exact-MSR codes for the same set of parameters. As compared with previous works, the major contributions of work include: i). An approach is proposed to directly include the d > 2k - 2 Exact-MSR codes in the product-matrix representation; ii). The required size of the finite field is reduced to n from n(d - k +1). The proposed coding techniques further improve the practicability of Exact-MSR codes.
  • Keywords
    codes; matrix algebra; distributed storage systems; erasure codes; exact-MSR codes; minimum storage regenerating code; product-matrix framework; Decoding; Encoding; Finite element analysis; Maintenance engineering; Symmetric matrices; Systematics; Vectors; Distributed storage; MSR codes; maximum-distance-separable(MDS) codes;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Personal Indoor and Mobile Radio Communications (PIMRC), 2013 IEEE 24th International Symposium on
  • Conference_Location
    London
  • ISSN
    2166-9570
  • Type

    conf

  • DOI
    10.1109/PIMRC.2013.6666251
  • Filename
    6666251