Paper 2024/1852 Original: eprint.iacr.org/2024/1852 ↗

Faster algorithms for isogeny computations over extensions of finite fields

Shiping Cai
Mingjie Chen
Christophe Petit
Abstract

Any isogeny between two supersingular elliptic curves can be defined over $\mathbb{F}_{p^2}$, however, this does not imply that computing such isogenies can be done with field operations in $\mathbb{F}_{p^2}$. In fact, the kernel generators of such isogenies are defined over extension fields of $\mathbb{F}_{p^2}$, generically with extension degree linear to the isogeny degree. Most algorithms related to isogeny computations are only efficient when the extension degree is small. This leads to efficient algorithms used in isogeny-based cryptographic constructions, but also limits their parameter choices at the same time. In this paper, we consider three computational subroutines regarding isogenies, focusing on cases with large extension degrees: computing a basis of $\ell$-torsion points, computing the kernel polynomial of an isogeny given a kernel generator, and computing the kernel generator of an isogeny given the corresponding quaternion ideal under the Deuring correspondence. We then apply our algorithms to the constructive Deuring correspondence algorithm from Eriksen, Panny, Sotáková and Veroni (LuCaNT'23) in the case of a generic prime characteristic, achieving around 30% speedup over their results.

Metadata
Available format(s)
PDF
Category
Public-key cryptography
Publication info
Preprint.
History
2024-11-12: last updated
2024-11-12: published
Short URL
https://ia.cr/2024/1852
License
https://creativecommons.org/licenses/by/4.0/

BibTeX

@misc{cryptoeprint:2024/1852,
      author = {Shiping Cai and Mingjie Chen and Christophe Petit},
      title = {Faster algorithms for isogeny computations over extensions of finite fields},
      howpublished = {Cryptology {ePrint} Archive, Paper 2024/1852},
      year = {2024},
      url = {https://eprint.iacr.org/2024/1852}
}
Note: In order to protect the privacy of readers, this archive does not use cookies or embedded third party content.