Secure Reusable Garbled Random Access Machine

碩士 === 國立中山大學 === 資訊工程學系研究所 === 103 === In 1980s, Yao proposed a secure computation which can allow two parties, one of which has a function f(⋅) and the other one has data x, to compute f(x) without revealing anything about f and x. Garbled circuits got numerous applications. But every time we want...

Full description

Bibliographic Details
Main Authors: Chih-Yu Chen, 陳志瑜
Other Authors: Chun-I Fan
Format: Others
Language:en_US
Published: 2015
Online Access:http://ndltd.ncl.edu.tw/handle/h47xfu
Description
Summary:碩士 === 國立中山大學 === 資訊工程學系研究所 === 103 === In 1980s, Yao proposed a secure computation which can allow two parties, one of which has a function f(⋅) and the other one has data x, to compute f(x) without revealing anything about f and x. Garbled circuits got numerous applications. But every time we want to garble a program, we need to convert it into a circuit first. In 2013, Lu and Ostrovsky proposed the garbled random-access machines, which can garble a RAM program directly. They constructed garbled RAM based on the pseudo-random functions and the Beaver-Micali-Rogaway paradigm in the semi-honest model. In Lu-Ostrovsky construction, the garbled data is reusable, but the garbled programs are not reusable. Therefore, we propose a reusable garbled RAM with both the reusable data and programs. Our first construction is designed in the semi-honest model for less computational cost. The second construction is based on the approach of Lindell and Pinkas, which used the cut-and-choose technique to resist the malicious adversary.