抗泄露可证明安全流密码研究

Side-channel attacks discovered by Kocher et al. in the late 90's, are cryptanalytic techniques that make use of physical leakages such as timing, power consumption, and can easily break many “mathematically sound” crypto-systems (e.g. AES, RSA) in practice. In FOCS 2008, Dziembowski and Pietrzak put forward the notion of leakage-resilient cryptography, and proposed a stream cipher provable secure against side-channel attacks, which is unprecedented in provable security where the primitives was usually treated as black-box. Later, Pietrzak simplified the construction and published it at Eurocrypt 2009, but the design still suffered from drawbacks such as key-size inefficiency and significant security loss. Subsequently, more researchers (including us) further simplified and improved the constructions in CCS 2010, CHES 2012, and CT-RSA 2013. In this paper, we provide an informative overview and introduction to all aforementioned constructions, including the improvements, drawbacks and the fundmental causes rooted in the designs. We also point out some mistakes in some public literatures, raise important open problems in this area, and look into the future directions of leakage-resilient cryptography.