Supersingular isogeny Diffie-Hellman (SIDH) is attractive for its relatively small public key size, but it is still unsatisfactory due to its efficiency, compared to other post-quantum proposals. In this paper, we focus on the performance of SIDH when the starting curve is E6 : y2 = x3 + 6x2 + x, which is fixed in Round-3 SIKE implementation. Inspired by previous works [1], [2], we present several tricks to accelerate key generation of SIDH and each process of SIKE. Our experimental results show that the performance of this work is at least 6.09% faster than that of the SIKE implementation, and we can further improve the performance when large storage is available.
Kaizhan LIN
Sun Yat-Sen University
Fangguo ZHANG
Sun Yat-sen University,Guangdong Key Laboratory of Information Security
Chang-An ZHAO
Sun Yat-Sen University,Guangdong Key Laboratory of Information Security
The copyright of the original papers published on this site belongs to IEICE. Unauthorized use of the original or translated papers is prohibited. See IEICE Provisions on Copyright for details.
Copy
Kaizhan LIN, Fangguo ZHANG, Chang-An ZHAO, "Faster Key Generation of Supersingular Isogeny Diffie-Hellman" in IEICE TRANSACTIONS on Fundamentals,
vol. E105-A, no. 12, pp. 1551-1558, December 2022, doi: 10.1587/transfun.2022EAP1026.
Abstract: Supersingular isogeny Diffie-Hellman (SIDH) is attractive for its relatively small public key size, but it is still unsatisfactory due to its efficiency, compared to other post-quantum proposals. In this paper, we focus on the performance of SIDH when the starting curve is E6 : y2 = x3 + 6x2 + x, which is fixed in Round-3 SIKE implementation. Inspired by previous works [1], [2], we present several tricks to accelerate key generation of SIDH and each process of SIKE. Our experimental results show that the performance of this work is at least 6.09% faster than that of the SIKE implementation, and we can further improve the performance when large storage is available.
URL: https://global.ieice.org/en_transactions/fundamentals/10.1587/transfun.2022EAP1026/_p
Copy
@ARTICLE{e105-a_12_1551,
author={Kaizhan LIN, Fangguo ZHANG, Chang-An ZHAO, },
journal={IEICE TRANSACTIONS on Fundamentals},
title={Faster Key Generation of Supersingular Isogeny Diffie-Hellman},
year={2022},
volume={E105-A},
number={12},
pages={1551-1558},
abstract={Supersingular isogeny Diffie-Hellman (SIDH) is attractive for its relatively small public key size, but it is still unsatisfactory due to its efficiency, compared to other post-quantum proposals. In this paper, we focus on the performance of SIDH when the starting curve is E6 : y2 = x3 + 6x2 + x, which is fixed in Round-3 SIKE implementation. Inspired by previous works [1], [2], we present several tricks to accelerate key generation of SIDH and each process of SIKE. Our experimental results show that the performance of this work is at least 6.09% faster than that of the SIKE implementation, and we can further improve the performance when large storage is available.},
keywords={},
doi={10.1587/transfun.2022EAP1026},
ISSN={1745-1337},
month={December},}
Copy
TY - JOUR
TI - Faster Key Generation of Supersingular Isogeny Diffie-Hellman
T2 - IEICE TRANSACTIONS on Fundamentals
SP - 1551
EP - 1558
AU - Kaizhan LIN
AU - Fangguo ZHANG
AU - Chang-An ZHAO
PY - 2022
DO - 10.1587/transfun.2022EAP1026
JO - IEICE TRANSACTIONS on Fundamentals
SN - 1745-1337
VL - E105-A
IS - 12
JA - IEICE TRANSACTIONS on Fundamentals
Y1 - December 2022
AB - Supersingular isogeny Diffie-Hellman (SIDH) is attractive for its relatively small public key size, but it is still unsatisfactory due to its efficiency, compared to other post-quantum proposals. In this paper, we focus on the performance of SIDH when the starting curve is E6 : y2 = x3 + 6x2 + x, which is fixed in Round-3 SIKE implementation. Inspired by previous works [1], [2], we present several tricks to accelerate key generation of SIDH and each process of SIKE. Our experimental results show that the performance of this work is at least 6.09% faster than that of the SIKE implementation, and we can further improve the performance when large storage is available.
ER -