Paper 2022/459

SIPFA: Statistical Ineffective Persistent Faults Analysis on Feistel Ciphers

Nasour Bagheri, Sadegh Sadeghi, Prasanna Ravi, Shivam Bhasin, and Hadi Soleimany


Persistent Fault Analysis (PFA) is an innovative and powerful analysis technique in which fault persists throughout the execution. The prior prominent results on PFA were on SPN block ciphers, and the security of Feistel ciphers against this attack has received less attention. In this paper, we introduce a framework to utilize Statistical Ineffective Fault Analysis (SIFA) in the persistent fault setting by proposing Statistical Ineffective Persistent Faults Analysis (SIPFA) that can be efficiently applied to Feistel ciphers in a variety of scenarios. To demonstrate the effectiveness of our technique, we apply SIFPA on three widely used Feistel schemes, DES, 3DES, and Camellia. Our analysis reveals that the secret key of these block ciphers can be extracted with a complexity of at most $2^{50}$ utilizing a single unknown fault. Furthermore, we demonstrate that the secret can be recovered in a fraction of a second by increasing the adversary's control over the injected faults. To evaluate SIPFA in a variety of scenarios, we conducted both simulations and real experiments utilizing electromagnetic fault injection on DES and 3DES.

Note: To appear at TCHES 2022-3

Available format(s)
Publication info
Published by the IACR in Tches 2022
Contact author(s)
na bagheri @ gmail com
sbhasin @ ntu edu sg
hadi soleimany @ gmail com
s sadeghi khu @ gmail com
prasanna ravi @ ntu edu sg
2022-04-12: last of 2 revisions
2022-04-12: received
See all versions
Short URL
Creative Commons Attribution


      author = {Nasour Bagheri and Sadegh Sadeghi and Prasanna Ravi and Shivam Bhasin and Hadi Soleimany},
      title = {SIPFA: Statistical Ineffective Persistent Faults Analysis on Feistel Ciphers},
      howpublished = {Cryptology ePrint Archive, Paper 2022/459},
      year = {2022},
      note = {\url{}},
      url = {}
Note: In order to protect the privacy of readers, does not use cookies or embedded third party content.