arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2302.02230cs.ITcs.CRmath.IT
收藏

具有最优下载速率和最优文件大小的$k$-服务器抗拜占庭PIR方案

$k$-server Byzantine-Resistant PIR Scheme with Optimal Download Rate and Optimal File Size

Stanislav Kruglik, Son Hoang Dau, Han Mao Kiah, Huaxiong Wang

AI总结:

针对$k$台可能串通或返回错误答案的服务器上的私有信息检索问题,提出一种文件大小较小的可达方案,证明该文件大小在固定检索符号数下最小,解决了相关公开问题,且方案达到最优下载速率。

摘要:

我们研究在$k$台服务器上复制的$m$个文件上设计私有信息检索(Private Information Retrieval,PIR)方案的问题,这些服务器可能串通,甚至更糟的是,可能返回错误答案。我们的目标是正确检索特定消息,同时对数据库服务器保密其身份。我们考虑该问题的渐近信息论容量,其定义为:当存储的文件数量足够大时,正确检索的符号数与下载符号数的最大比值。我们提出了一种文件大小较小的可达方案,并证明对于固定的检索符号数,这样的文件大小是最小的,解决了Banawan和Ulukus提出的问题。

英文摘要:

We consider the problem of designing a Private Information Retrieval (PIR) scheme on $m$ files replicated on $k$ servers that can collude or, even worse, can return incorrect answers. Our goal is to correctly retrieve a specific message while keeping its identity private from the database servers. We consider the asymptotic information-theoretic capacity of this problem defined as the maximum ratio of the number of correctly retrieved symbols to the downloaded one for a large enough number of stored files. We propose an achievable scheme with a small file size and prove that such a file size is minimal for the fixed number of retrieved symbols, solving the problem pointed out by Banawan and Ulukus.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑