期刊文献+

去中心化存储下分布式低带宽多节点修复方法 被引量:1

Distributed Low Bandwidth Multi-node Repair Under Decentralized Storage
下载PDF
导出
摘要 为扩展纠删码在区块链中的应用,研究了去中心化存储系统中的修复机制,发现系统中RS码存在多节点修复成本高、效率低的问题.针对这个问题,本文提出一种更契合去中心化网络环境下的多节点修复传输模型DSMR,充分利用RS码在修复多节点时的数据冗余性和计算冗余性.通过节点稳定性和网络跳数来选择节点、构造数据传输并行结构、分组修复计算和节点数据交互四个步骤,以最大化数据传输效率、较低带宽开销和较短修复时间完成了去中心下分布式低带宽多节点修复工作.理论及实验结果表明,在任意(n,k)参数下,分布式多节点修复传输机制在保持较低存储空间地同时进一步降低修复带宽、减少修复时间. In order to extend the application of erasure codes in block chain,the repair mechanism in decentralized storage system is studied,found that RS(Reed-Solomon)in the system have problems of high cost and low efficiency.Aiming at this problem,This paper proposes a multi-node repair transmission model DSMR(Distributed Structure Based Multinode Repair)which is more suitable for decentralized network environment,making full use of data redundancy and computational redundancy when RS code is used to repair multiple nodes.Through the four steps of selecting nodes through node stability and network hops,constructing the parallel structure of data transmission,grouping repair calculation and node data interaction,the distributed multi-node repair with low bandwidth under decentralization was completed in order to maximize the data transmission efficiency,lower bandw idth overhead and shorter repair time.Theoretical and experimental results show that under any(n,k)parameter,the distributed multi-node repair transmission mechanism further reduces the repair bandw idth and the repair time while maintaining a low er storage space.
作者 李慧 李贵洋 周悦 江小玉 韩鸿宇 LI Hui;LI Gui-yang;ZHOU Yue;JIANG Xiao-yu;HAN Hong-yu(Department of Computer Science,Sichuan Normal University,Chengdu 610101,China)
出处 《小型微型计算机系统》 CSCD 北大核心 2020年第7期1553-1558,共6页 Journal of Chinese Computer Systems
基金 国家自然科学基金项目(61701331)资助。
关键词 去中心化存储 纠删码 多节点修复 数据传输 网络带宽 decentralized storage erasure codes multi-node repair data transmission network bandwidth
  • 相关文献

参考文献6

二级参考文献62

共引文献2866

同被引文献11

引证文献1

二级引证文献1

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部