摘要
基于格的加密算法是可以抵抗量子攻击的后量子密码的一个重要候选,而基于错误学习的加密算法是具有强安全性证明的格加密算法。文中在GPU和CPU上实现了多比特错误学习加密算法。实验结果表明,除了在GPU得到了硬件加速比外,基于错误学习的加密算法的加解密时间结果使得该算法具有更好的实用性。
Lattice-based cryptosystem is one of the important candidates for post-quantum cbyptography and cryptosystem based on learning with errors is the most efficient in lattice-based cryptosystems with strong security proof. We present a practical implementation of LWE-based cryptosystem both on CPU and GPU. The experiment shows that besides the hardware speedup of GPU, the LWE-based cryptosystem has a more practical encryption and better decryption time.
出处
《电子科技》
2013年第4期93-95,共3页
Electronic Science and Technology
关键词
错误学习
格加密
后量子密码
GPU
learning with errors
lattice-based cryptography
post-quantum cryptography
GPU
