Paper 2016/526

Extended Tower Number Field Sieve with Application to Finite Fields of Arbitrary Composite Extension Degree

Taechan Kim and Jinhyuck Jeong

Abstract

We propose a generalization of exTNFS algorithm recently introduced by Kim and Barbulescu (CRYPTO 2016). The algorithm, exTNFS, is a state-of-the-art algorithm for discrete logarithm in Fpn in the medium prime case, but it only applies when n=ηκ is a composite with nontrivial factors η and κ such that gcd(η,κ)=1. Our generalization, however, shows that exTNFS algorithm can be also adapted to the setting with an arbitrary composite n maintaining its best asymptotic complexity. We show that one can solve discrete logarithm in medium case in the running time of Lpn(1/3,48/93) (resp. Lpn(1/3,1.71) if multiple number fields are used), where n is an \textit{arbitrary composite}. This should be compared with a recent variant by Sarkar and Singh (Asiacrypt 2016) that has the fastest running time of Lpn(1/3,64/93) (resp. Lpn(1/3,1.88)) when is a power of prime 2. When is of special form, the complexity is further reduced to . On the practical side, we emphasize that the keysize of pairing-based cryptosystems should be updated following to our algorithm if the embedding degree remains composite.

Metadata
Available format(s)
PDF
Publication info
Preprint. MINOR revision.
Keywords
Discrete Logarithm ProblemNumber Field SieveFinite FieldsCryptanalysis
Contact author(s)
yoshiki1 @ snu ac kr
History
2016-10-06: revised
2016-05-29: received
See all versions
Short URL
https://ia.cr/2016/526
License
Creative Commons Attribution
CC BY

BibTeX

@misc{cryptoeprint:2016/526,
      author = {Taechan Kim and Jinhyuck Jeong},
      title = {Extended Tower Number Field Sieve with Application to Finite Fields of Arbitrary Composite Extension Degree},
      howpublished = {Cryptology {ePrint} Archive, Paper 2016/526},
      year = {2016},
      url = {https://eprint.iacr.org/2016/526}
}
Note: In order to protect the privacy of readers, eprint.iacr.org does not use cookies or embedded third party content.