嵌入式系统芯片中SM2算法软硬件协同设计与实现

发布时间：2019-06-22 13:02

【摘要】：针对现有的椭圆曲线算法系统级设计中开发周期长,以及不同模块的性能开销指标不明确等问题,提出一种基于电子系统级(ESL)设计的软硬件(HW/SW)协同设计方法。该方法通过分析SM2(Shang Mi2)算法原理与实现方式,研究了不同的软硬件划分方案,并采用统一建模语言System C对硬件模块进行周期精确级建模。通过模块级与系统级两层验证比较软硬件模块执行周期数,得出最佳性能划分方式。最后结合算法控制流程图(CFG)与数据流程图(DFG)将ESL模型转化为寄存器传输级(RTL)模型进行逻辑综合与比较,得出在180 nm CMOS工艺,50 MHz频率下,当算法性能最佳时,点乘模块执行时间为20 ms,门数83 000,功耗约2.23 m W。实验结果表明所提系统级架构分析对基于椭圆曲线类加密芯片在性能、面积与功耗的评估优势明显且适用性强,基于此算法的嵌入式系统芯片(So C)可根据性能与资源限制选择合适的结构并加以应用。
[Abstract]:In order to solve the problems of long development cycle and unclear performance overhead of different modules in the existing Elliptic Curve algorithm (Elliptic Curve algorithm) system level design, a software and hardware (HW/SW) co-design method based on electronic system level (ESL) design is proposed. By analyzing the principle and implementation of SM2 (Shang Mi2) algorithm, different software and hardware partition schemes are studied, and System C is used to model the hardware module at the periodic and accurate level. Through the verification and comparison of the execution cycle of software and hardware modules at the module level and the system level, the optimal performance partition method is obtained. Finally, the ESL model is transformed into register transfer level (RTL) model by combining the algorithm control flow chart (CFG) and the data flow chart (DFG) to synthesize and compare the logic. It is concluded that when the algorithm performance is the best, the point multiplication module execution time is 20 ms, gate number 83 000, and the power consumption is about 2.23 MW at 180 nm CMOS process and 50 MHz frequency. The experimental results show that the proposed system-level architecture analysis has obvious advantages and strong applicability in evaluating the performance, area and power consumption of the Elliptic Curve encryption chip. The embedded system chip (So C) based on this algorithm can select the appropriate structure according to the performance and resource constraints and apply it.
【作者单位】：湖南大学物理与微电子科学学院;湖南大学信息科学与工程学院;
【基金】：国家自然科学基金资助项目(61300037)
【分类号】：TP309;TP368.1

【参考文献】