Formal models for quantitative analysis of software security
Japan Society for the Promotion of Science:Grants-in-Aid for Scientific Research
Date (from‐to) : 2014/04 -2018/03
Author : Seki Hiroyuki; HASHIMOTO KENJI
A few quantitative notions for security and privacy of software such as quantitative information flow (QIF) and differential privacy have been proposed. In this research, we developed methods that analyze given programs or systems based on such notions. Specifically, we proposed an approximation algorithm that computes leakage by timing attack against an RSA decoder, a verification algorithm of k-secrecy of XML databases. Furthermore, as a theoretical basis for QIF analysis of programs that dynamically generate strings, we propose algorithms that counts, for a given recognizable or algebraic series S and a natural number d, the summation of the coefficients (or weights) of words of length d in S efficiently. The proposed methods were shown to be effective either by computer simulation or by experiments based on the implemented tools.