Paper 2017/415

Towards Practical PFE: An Efficient 2-Party Private Function Evaluation Protocol Based on Half Gates

Osman Bicer, Muhammed Ali Bingol, Mehmet Sabir Kiraz, and Albert Levi

Abstract

Private function evaluation (PFE) is a special case of secure multi-party computation (MPC), where the function to be computed is known by only one party. PFE is useful in several real-life settings where an algorithm or a function itself needs to remain secret due to its confidential classification or intellectual property. In this work, we look back at the seminal PFE framework presented by Mohassel and Sadeghian at Eurocrypt’13. We show how to adapt and utilize the well-known half gates garbling technique (Zahur et al., Eurocrypt’15) to their constant round 2-party PFE scheme. Compared to their scheme, our resulting optimization considerably improves the efficiency of both the underlying Oblivious Evaluation of Extended Permutation (OEP) and secure 2-party computation (2PC) protocol, and yields a more than 40% reduction in overall communication cost (the computation time is also slightly decreased, and the number of rounds remains unchanged).

Metadata
Available format(s)
PDF
Category
Cryptographic protocols
Publication info
Preprint.
Keywords
Private function evaluationSecure 2-party computationCommunication complexityCryptographic protocol.
Contact author(s)
osmnbicr @ gmail com
History
2020-07-14: revised
2017-05-15: received
See all versions
Short URL
https://ia.cr/2017/415
License
Creative Commons Attribution
CC BY

BibTeX

@misc{cryptoeprint:2017/415,
      author = {Osman Bicer and Muhammed Ali Bingol and Mehmet Sabir Kiraz and Albert Levi},
      title = {Towards Practical PFE: An Efficient 2-Party Private Function Evaluation Protocol Based on Half Gates},
      howpublished = {Cryptology ePrint Archive, Paper 2017/415},
      year = {2017},
      note = {\url{https://eprint.iacr.org/2017/415}},
      url = {https://eprint.iacr.org/2017/415}
}
Note: In order to protect the privacy of readers, eprint.iacr.org does not use cookies or embedded third party content.