面向序列密码的高效能分层式比特抽取网络设计研究

摘要

针对序列密码算法中抽取操作的可重构硬件实现资源消耗大的问题,通过研究序列密码中非线性布尔函数的变量需求,基于Inverse Butterfly网络提出一种高效能分层式比特抽取网络-HHBN(High-efficiency hierar-chical bit-extraction network).与其他网络进行对比,该网络可一次抽取出含有重复变量的多组数据,且不仅其实际性能与灵活度优于其他大多网络,面积消耗也远小于同灵活度的Crossbar网络.在Synopsys公司的Design Compiler进行了综合,实验结果表明与Crossbar网络在不同位宽实现相比,其面积减少约20％ ～ 50％,这就减少了实现不同序列密码算法抽取操作的可重构硬件资源消耗,从而提高了效能.%Aiming at the problems in stream cipher algorithm that the high areas of the reconfigurable hardware implementation of bit-extraction,we propose a high-efficiency hierarchical bit-extraction network based on the Inverse Butterfly network named HHBN (High-efficiency hierarchical bit-extraction network),through the studying of the nonlineal boolean function in stream cipher algorithm.Compared with other networks,the proposed network can extract multiple sets data with common variables.The actual performance and flexibility of HHBN are better than most other networks,and area is far less than the Crossbar network that has same flexibility with HHBN.We integrated the design compiler software at Synopsys company.The results show that compared with the Crossbar network in different width,HHBN reduce the area about 20％ ～ 50％,which reduces the resources consumption of extraction operation for different stream cipher algorithm,so as to improve the efficiency of it.