基于CRCLA的形式化描述与退火遗传映射算法  

Formal description and annealing genetic mapping algorithm based on CRCLA

作　　者：杨晨光 李伟[1] 杜怡然 陈韬[1] YANG Chen-guang;LI Wei;DU Yi-ran;CHEN Tao(School of Cryptography Engineering,Strategic Support Forces Information Engineering University,Zhengzhou 450001,China)

机构地区：[1]战略支援部队信息工程大学密码工程学院,河南郑州450001

出　　处：《计算机工程与设计》2024年第5期1368-1375,共8页Computer Engineering and Design

基　　金：国家自然科学基金项目(61404175)。

摘　　要：为解决密码算法映射到粗粒度可重构密码逻辑阵列(CRCLA)过程的描述困难问题,提出一种数据流图节点与划分后节点簇的描述形式,并以一种坐标序列描述形式精确显示CRCLA中资源的占用情况;针对密码算法映射到CRCLA上性能不高及编译时间过长的问题,提出一种能快速收敛到全局最优解的退火遗传算法,从初始温度、降温系数等方面对退火算法进行改进,增加遗传算法的筛选、交叉与变异流程。实验结果表明,编译时间与性能平均降低了30.6%与13.4%,验证了算法的高能效映射。To solve the difficulty of describing the process of mapping cryptographic algorithms to coarse-grained reconfigurable cryptographic logical array(CRCLA),a description form of data flow graph nodes and node clusters after partition was proposed,and a coordinate sequence description form was used to accurately display the resource usage in CRCLA.To solve the problems of poor performance and long compiling time of the cryptographic algorithm on CRCLA,an annealing genetic algorithm which could converge to the global optimal solution quickly was proposed.The annealing algorithm was improved from the aspects of initial temperature and cooling coefficient,and the screening,crossing and mutation process of the genetic algorithm was increased.Experimental results show that the compilation time and performance are reduced by 30.6%and 13.4%on ave-rage,which verifies the high energy efficiency mapping of the algorithm.

关 键 词：密码算法 数据流图 划分 描述形式 粗粒度可重构密码逻辑阵列 快速收敛 退火遗传算法 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]