首页 | 本学科首页   官方微博 | 高级检索  
     检索      

基于单元分类的CMOL常开缺陷单元容错映射
引用本文:顾贤贵,夏银水.基于单元分类的CMOL常开缺陷单元容错映射[J].宁波大学学报(理工版),2020,33(1):51-57.
作者姓名:顾贤贵  夏银水
作者单位:宁波大学 信息科学与工程学院, 浙江 宁波 315211
摘    要:CMOS纳米分子混合电路(CMOS/nanowire/MOLeclular hybrid circuits, CMOL)在制造过程中会引入较高缺陷率, 从而导致可用映射资源的减少. 针对由此产生的映射困难问题, 本文采用单元分类思想, 对部分缺陷单元加以利用, 以增加可映射单元数, 进而提高映射成功率. 首先根据单元缺陷类型的差异, 将缺陷单元分为可用和不可用两类进行标记, 然后对可用缺陷单元加以利用, 并采用改进的进化算法完成单元容错映射. 实验结果表明, 与已有方法相比, 新方法在运行效率和成功率上分别得到了19.17%和30.14%的提升.

关 键 词:CMOL  进化算法  单元容错映射

Cell classification based CMOL stuck-at-open defect-tolerant cell mapping
GU Xiangui,XIA Yinshui.Cell classification based CMOL stuck-at-open defect-tolerant cell mapping[J].Journal of Ningbo University(Natural Science and Engineering Edition),2020,33(1):51-57.
Authors:GU Xiangui  XIA Yinshui
Institution:Faculty of Electrical Engineering and Computer Science, Ningbo University, Ningbo 315211, China
Abstract:To tackle the problem of high defect rate in CMOS/nanowire/MOLeclular hybrid circuits (CMOL), a cell classification based defect-tolerant cell mapping method is proposed. First, recoverable and unrecoverable defective cells are marked, and the recoverable defective ones are put to use again. Then, a modified evolutionary algorithm is used to complete the defect-tolerant cell mapping. Compared with the existing approaches, the CPU time and success rate using the proposed method are optimized to 19.17% and 30.14%, respectively.
Keywords:CMOL  evolutionary algorithm  defect-tolerant cell mapping
本文献已被 CNKI 万方数据 等数据库收录!
点击此处可从《宁波大学学报(理工版)》浏览原始摘要信息
点击此处可从《宁波大学学报(理工版)》下载免费的PDF全文
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号