上海交通大学学报 ›› 2024, Vol. 58 ›› Issue (10): 1596-1605.doi: 10.16183/j.cnki.jsjtu.2023.151

• 电子信息与电气工程 • 上一篇    下一篇

三角形结合方案的最优局部修复码构造

王静(), 李静辉, 杨佳蓉, 王娥   

  1. 长安大学 信息工程学院,西安 710018
  • 收稿日期:2023-04-21 修回日期:2023-09-22 接受日期:2023-09-25 出版日期:2024-10-28 发布日期:2024-11-01
  • 作者简介:王 静(1982—),博士,教授,从事网络编码及分布式存储编码等方面的研究; E-mail:jingwang@chd.edu.cn.
  • 基金资助:
    国家自然科学基金(62001059);陕西省自然科学基金资助项目(2022JM-056);长安大学大学生创新创业训练计划项目(S202310710121)

Construction of Optimal Locally Repairable Codes of Triangular Association Schemes

WANG Jing(), LI Jinghui, YANG Jiarong, WANG E   

  1. School of Information Engineering, Chang’an University, Xi’an 710018, China
  • Received:2023-04-21 Revised:2023-09-22 Accepted:2023-09-25 Online:2024-10-28 Published:2024-11-01

摘要:

局部修复码(LRCs)为用于分布式存储系统中的新型纠删码,能够有效实现海量数据的可靠高效存储,构造具有(r, t)局部性的LRCs已成为当前研究热点.为此,提出基于三角形结合方案的LRCs构造方法,可构造具有任意(r, t)局部性的二元最优LRCs.性能分析结果表明,构造的可用性t=2的LRCs达到了最优码率界,构造的具有任意局部性r>2和可用性t>2的LRCs达到了最优最小距离界.与基于近正则图及基于直积码等构造方法相比,本文构造出的LRCs在码率上表现更优且参数选择更灵活.

关键词: 分布式存储系统, 局部修复码, 三角形结合方案, 最小距离

Abstract:

As a new erasure code for distributed storage systems, locally repairable codes (LRCs) can effectively realize the reliable and efficient storage of massive data. The construction of locally repairable codes with (r,t) locality has become a research hotspot recently. Therefore, the construction methods of locally repairable codes based on triangular association schemes are proposed, which can construct optimal binary locally repairable codes with arbitrary (r,t) locality. Performance analyses show that the LRCs constructed with availability t=2 reach the optimal code rate bound, the LRCs constructed with arbitrary locality r>2 and availability t>2 reach the optimal minimum distance bound. The LRC constructed in this paper performs better in terms of code rate and more flexible parameter selection than those constructed based on near-regular graphs and direct product codes, etc.

Key words: distributed storage system, locally repairable code (LRC), triangular association schemes, minimum distance

中图分类号: