粗粒度可重构密码逻辑阵列智能映射算法研究

针对粗粒度可重构密码逻辑阵列密码算法映射周期长且性能不高的问题，该文通过构建粗粒度可重构密码逻辑阵列参数化模型，以密码算法映射时间及实现性能为目标，结合本文构建的粗粒度可重构密码逻辑阵列结构特征，提出了一种算法数据流图划分算法.通过将密码算法数据流图中节点聚集成簇并以簇为最小映射粒度进行映射，降低算法映射复杂度；该文借鉴机器学习过程，构建了具备学习能力的智慧蚁群模型，提出了智慧蚁群优化算法，通过对训练样本的映射学习，持续优化初始化信息素浓度矩阵，提升算法映射收敛速度，以已知算法映射指导未知算法映射，实现密码算法映射的智能化.实验结果表明，本文提出的映射方法能够平均降低编译时间37.9%并实现密码算法映射性能最大，同时，以算法数据流图作为映射输入，自动化的生成密码算法映射流，提升了密码算法映射的直观性与便捷性.

Research on Coarse-Grained Reconfigurable Cryptographic Logic Array Intelligent Mapping Algorithm

This paper presents algorithms for cryptographic algorithms mapping based on the coarse-grained reconfigurable cryptographic logic array.Due to a long mapping period and low performance for current algorithms mapping,we propose two methods to improve it.First,combine with the structural characteristics of the coarse-grained reconfigurable cryptographic logic array and cryptographic algorithms,an algorithm for data flow graph partitioning is proposed.By integrating the nodes into clusters to reduce the mapping complexity.Second,refer to the idea of machine learning,a smart ant colony optimization algorithm is proposed.By learning the training samples,the pheromone concentration matrix is continuously optimized and realizes the intelligentization of cryptographic algorithm mapping.The experimental results show that the proposed mapping method can reduce the compilation time by 37.9% and achieve the best performance.At the same time,the algorithm data flow graph is used as the mapping input,and the cryptographic algorithm map stream is automatically generated,which improves the cryptographic algorithm mapping more intuitive and convenient.