Proving differential privacy in Hoare logic

被引:33
|
作者
Barthe, Gilles [1 ]
Gaboardi, Marco [2 ]
Arias, Emilio Jesus Gallego [3 ]
Hsu, Justin [3 ]
Kunz, Cesar [1 ]
Strub, Pierre-Yves [1 ]
机构
[1] IMDEA Software Inst, Madrid, Spain
[2] Univ Dundee, Dundee DD1 4HN, Scotland
[3] Univ Penn, Philadelphia, PA 19104 USA
关键词
SECURE INFORMATION-FLOW;
D O I
10.1109/CSF.2014.36
中图分类号
TP301 [理论、方法];
学科分类号
081202 ;
摘要
Differential privacy is a rigorous, worst-case notion of privacy-preserving computation. Informally, a probabilistic program is differentially private if the participation of a single individual in the input database has a limited effect on the program's distribution on outputs. More technically, differential privacy is a quantitative 2-safety property that bounds the distance between the output distributions of a probabilistic program on adjacent inputs. Like many 2-safety properties, differential privacy lies outside the scope of traditional verification techniques. Existing approaches to enforce privacy are based on intricate, non-conventional type systems, or customized relational logics. These approaches are difficult to implement and often cumbersome to use. We present an alternative approach that verifies differential privacy by standard, non-relational reasoning on non-probabilistic programs. Our approach transforms a probabilistic program into a non-probabilistic program which simulates two executions of the original program. We prove that if the target program is correct with respect to a Hoare specification, then the original probabilistic program is differentially private. We provide a variety of examples from the differential privacy literature to demonstrate the utility of our approach. Finally, we compare our approach with existing verification techniques for privacy.
引用
收藏
页码:411 / 424
页数:14
相关论文
共 50 条
  • [1] Proving pointer programs in Hoare logic
    Bornat, R
    [J]. MATHEMATICS OF PROGRAM CONSTRUCTION, 2000, 1837 : 102 - 126
  • [2] PROVING PROGRAM INCLUSION USING HOARE LOGIC
    BERGSTRA, JA
    KLOP, JW
    [J]. THEORETICAL COMPUTER SCIENCE, 1984, 30 (01) : 1 - 48
  • [3] Hyper Hoare Logic: (Dis-)Proving Program Hyperproperties
    Dardinier, Thibault
    Muller, Peter
    [J]. PROCEEDINGS OF THE ACM ON PROGRAMMING LANGUAGES-PACMPL, 2024, 8 (PLDI):
  • [4] Proving Differential Privacy with Shadow Execution
    Wang, Yuxin
    Ding, Zeyu
    Wang, Guanhong
    Kifer, Daniel
    Zhang, Danfeng
    [J]. PROCEEDINGS OF THE 40TH ACM SIGPLAN CONFERENCE ON PROGRAMMING LANGUAGE DESIGN AND IMPLEMENTATION (PLDI '19), 2019, : 655 - 669
  • [5] Proving Differential Privacy via Probabilistic Couplings
    Barthe, Gilles
    Gaboardi, Marco
    Gregoire, Benjamin
    Hsu, Justin
    Strub, Pierre-Yves
    [J]. PROCEEDINGS OF THE 31ST ANNUAL ACM-IEEE SYMPOSIUM ON LOGIC IN COMPUTER SCIENCE (LICS 2016), 2016, : 749 - 758
  • [6] Reverse Hoare Logic
    de Vries, Edsko
    Kontavas, Vasileios
    [J]. SOFTWARE ENGINEERING AND FORMAL METHODS, 2011, 7041 : 155 - 171
  • [7] Hoare logic in the abstract
    Martin, Ursula
    Mathiesen, Erik A.
    Oliva, Paulo
    [J]. COMPUTER SCIENCE LOGIC, PROCEEDINGS, 2006, 4207 : 501 - 515
  • [8] THE HOARE LOGIC OF CSP, AND ALL THAT
    LAMPORT, L
    SCHNEIDER, FB
    [J]. ACM TRANSACTIONS ON PROGRAMMING LANGUAGES AND SYSTEMS, 1984, 6 (02): : 281 - 296
  • [9] HOARE LOGIC, EXECUTABLE SPECIFICATIONS, AND LOGIC PROGRAMS
    FUCHS, NE
    [J]. STRUCTURED PROGRAMMING, 1992, 13 (03): : 129 - 135
  • [10] Matching Logic: An Alternative to Hoare/Floyd Logic
    Rosu, Grigore
    Ellison, Chucky
    Schulte, Wolfram
    [J]. ALGEBRAIC METHODOLOGY AND SOFTWARE TECHNOLOGY, 2011, 6486 : 142 - +