SKINNY is a family of tweakable lightweight block ciphers, proposed in CRYPTO 2016. The proposal of SKINNY describes two block size variants of 64 and 128 bits as well as three options for tweakey. In this paper, we present fault attacks (FA) on all SKINNY variants. In the first part of the paper, we propose differential fault analysis (DFA) attacks on SKINNY variants keeping the tweak fixed. The attack model of tweakable block ciphers allows the access and full control of the tweak by the attacker. Respecting this attack model, we assume a fixed tweak for the attack window. With this assumption, extraction of the master key of SKINNY requires about 10 random nibble fault injections on average for 64-bit versions of the cipher, whereas the 128-bit versions require roughly 21 byte-fault-injections. In the later part of this work, we relax this assumption and perform fault attacks under known but randomly varying tweaks. It is found that pairs of bit faults at the input and output of the S-Boxes allow complete key recovery under random tweak. Moreover, explicit access to ciphertexts is not required in our attack, and key recovery is possible only by knowing if the ciphertext is correct or faulty. This property of the attack allows key recovery even at the presence of simple redundancy-based FA countermeasures. Both the DFA and paired fault-based attacks were validated through extensive simulation. To the best of authors' knowledge, these are the first instances of FAs reported on SKINNY tweakable block cipher family.
In order to obtain the secret key, the majority of physical attacks require knowledge of the plaintext or ciphertext, which may be unavailable or cannot be exploited. Blind attacks are introduced to do key recovery in circumstances where the adversary has no direct access to plaintext and ciphertext. A combination of fault and power attacks can circumvent typical countermeasures in this setting, for example, Fault Template Attack (FTA). However, FTA relies on bit fault injection, which is difficult to implement in practice. The SIFA-blind, a framework for executing the Statistical Ineffective Fault Attack, is more flexible, but sensitivity to setup noise and missed faults is its main drawback. To address this deficiency, we suggest two ways to use Statistical Effective Fault Attack in a blind setting that are much less affected by missed faults and noise when measuring power traces, even though they do not use fault injection at the bit level. In order to demonstrate the viability and adaptability of our proposed attacks, we injected a fault via glitch frequency onto the ChipWhisperer board. While SEFA-blind does not need a bit-level fault, our results demonstrate that it is better than SIFA-blind when the number of missed faults increases.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.