Paper 2020/1061

A new idea in response to fast correlation attacks on small-state stream ciphers

Vahid Amin Ghafari and Fujiang Lin

Abstract

In the conference “Fast Software Encryption 2015”, a new line of research was proposed by introducing the first small-state stream cipher (SSC). The goal was to design lightweight stream ciphers for hardware application by going beyond the rule that the internal state size must be at least twice the intended security level. Time-memory-data trade-off (TMDTO) attacks and fast correlation attacks (FCA) were successfully applied to all proposed SSCs which can be implemented by less than 1000 gate equivalents in hardware. It is possible to increase the security of stream ciphers against FCA by exploiting more complicated functions for the nonlinear feedback shift register and the output function, but we use lightweight functions to design the lightest SSC in the world while providing more security against FCA. Our proposed cipher provides 80-bit security against TMDTO distinguishing attacks, while Lizard and Plantlet provide only 60-bit and 58-bit security against distinguishing attacks, respectively. Our main contribution is to propose a lightweight round key function with a very long period that increases the security of SSCs against FCA.

Metadata
Available format(s)
PDF
Category
Secret-key cryptography
Publication info
Preprint. MINOR revision.
Keywords
Stream ciphersLightweight cipherGrainHardware implementationCryptographic primitive
Contact author(s)
vahidaming @ yahoo com
History
2020-11-08: revised
2020-09-03: received
See all versions
Short URL
https://ia.cr/2020/1061
License
Creative Commons Attribution
CC BY

BibTeX

@misc{cryptoeprint:2020/1061,
      author = {Vahid Amin Ghafari and Fujiang Lin},
      title = {A new idea in response to fast correlation attacks on small-state stream ciphers},
      howpublished = {Cryptology {ePrint} Archive, Paper 2020/1061},
      year = {2020},
      url = {https://eprint.iacr.org/2020/1061}
}
Note: In order to protect the privacy of readers, eprint.iacr.org does not use cookies or embedded third party content.