跳到主要导航 跳到搜索 跳到主要内容

Research on data recovery technology based on flash memory device

  • Lele Guan*
  • , Jun Zheng
  • , Chenyang Li
  • , Dianxin Wang
  • *此作品的通讯作者
  • Beijing Institute of Technology

科研成果: 书/报告/会议事项章节 › 会议稿件 › 同行评审

摘要

Due to the significant internal structural difference between flash memory devices and traditional mechanical hard disks, the data recovery technology for traditional mechanical hard disks cannot be directly applied to flash memory devices. Therefore, there is an urgent need to explore a data recovery method specific to flash storage devices. On the premise of obtaining the mirror data structure parameters of the flash memory chip, through analyzing the underlying data from the said chip, determination and analysis can be done to the interleave type of data on the physical structure. And according to the different interleaving granularity, universal formulas can be generalized one by one. Select the data of valid block to be put into the formula which is suitable for interleaving granularity, and then calculate the corresponding logical address. Finally, write the buffers one by one in the order indicated by the logical address to achieve the data recovery. Under the circumstance where the structure parameters cannot be obtained due to mechanical malfunction, to develop a recovery strategy based on an unknown algorithm by constructing a learning disk can be both practical and instructive.

源语言英语
主期刊名Algorithms and Architectures for Parallel Processing - 18th International Conference, ICA3PP 2018, Proceedings
编辑Jaideep Vaidya, Jin Li
出版商Springer Verlag
页263-271
页数9
ISBN(印刷版)9783030050535
DOI
出版状态已出版 - 2018
活动18th International Conference on Algorithms and Architectures for Parallel Processing, ICA3PP 2018 - Guangzhou, 中国
期限: 15 11月 2018 → 17 11月 2018

丛书

姓名Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
卷11335 LNCS
ISSN(印刷版)0302-9743
ISSN(电子版)1611-3349

会议

会议18th International Conference on Algorithms and Architectures for Parallel Processing, ICA3PP 2018
国家/地区中国
市Guangzhou
时期15/11/18 → 17/11/18

学术指纹

探究 'Research on data recovery technology based on flash memory device' 的科研主题。它们共同构成独一无二的学术指纹。

引用此