Destroying Fault Invariant with Randomization

Abstract

Researchers have demonstrated the ineffectiveness of deterministic countermeasures and emphasized on the use of randomness for protecting cryptosystems against fault attacks. One such countermeasure for AES was proposed in LatinCrypt 2012, which masks the faulty output with secret values. However this countermeasure does not affect the erroneous byte in the faulty computation of the last AES round and is thus shown to be flawed in FDTC 2013. In this paper, we examine the Latin-Crypt 2012 countermeasure in detail and identify its additional flaws in order to develop a robust countermeasure. We bring out the major weakness in the infection mechanism of the LatinCrypt 2012 countermeasure which not only makes the attack of FDTC 2013 much more flexible, but also enables us to break this seemingly complex countermeasure using Piret & Quisquater's attack that requires only 8 pairs of correct and faulty ciphertexts. Finally, we combine all our observations and propose a countermeasure that employs randomness much more effectively to prevent state-of-the-art differential fault attacks against AES.

Fault Tolerant Infective Countermeasure for AES

“…Hence, such a fault model is a realistic threat to embedded applications. We demonstrate in this paper that the instruction skip fault model weakens the infective countermeasure scheme proposed in [11] and allows easy key recovery. Thus, it is important to make infective countermeasure tolerant against such attacks.…”

“…4. The paper also presents experimental results to validate the enhanced security of the aforementioned countermeasure schemes in comparison with the original infective countermeasure scheme presented in [11]. We present both simulations and real-life experiments on a SASEBO-W platform to compare the fault resistance of each variety of the countermeasure.…”

“…The countermeasure scheme proposed by Tupsamudre et al [11] provides no formal guarantees of security against DFA. Moreover, fault attacks that allow an adversary to change the flow sequence of an algorithm by methods such as instruction skips have also not been considered in the security analysis of this countermeasure.…”

“…The paper presents a formal information theoretic proof of security for the infective countermeasure scheme proposed by Tupsamudre et al [11] against single and multiple fault injection models, under the assumption that an adversary cannot change the flow sequence or skip instructions. In particular, the paper analyzes the security of the countermeasure against both DFA and DFIA-like attacks.…”

“…A random variation of the infective countermeasure was proposed by Gierlichs et al [9]. However, the infection method employed by this countermeasure has a number of shortcomings, as demonstrated by Battistello and Giraud [10], and in greater detail by Tupsamudre et al [11]. Tupsamudre et al have also proposed an improved infective countermeasure that avoids all the pitfalls of [9] and is claimed to thwart DFA.…”