Abstract: Aiming to improve the computational performance of the domestic cryptographic algorithm SM9, this paper proposes a twodimensional Comb fixedbase modular exponentiation algorithm, extends the application of precomputed scalar multiplication, and optimizes the commonly used ID. Theoretical analysis and experimental tests show that these proposed methods can effectively improve the computational performance of components such as fixedbase modular exponentiation and the three common steps of the SM9 algorithm by precomputing and increasing acceptable storage overhead. After applying the above improvement methods comprehensively, the performance of SM9 digital signature generation and verification, key exchange, key encapsulation, and encryption algorithms is improved by 14% to 116%.

Key words: domestic cryptographic algorithm, SM9, computational performance, precomputation, Comb fixedbase

摘要： 针对国密算法SM9的计算性能改进问题，提出2维Comb固定基模幂算法、预计算标量乘的拓展应用、针对常用ID优化等计算性能改进方法，理论分析和实验测试表明，所提方法通过预计算并增加可以接受的存储开销，能有效提升固定底数模幂、SM9算法3个常见步骤等组件的计算性能.综合运用上述改进方法后，SM9数字签名的生成与验证、密钥交换、密钥封装、加密5项算法的性能提升幅度为14%~116%.

关键词: 国密算法, SM9, 计算性能, 预计算, Comb固定基