期刊文献+

实时SAR成像系统中矩阵转置的设计和实现 被引量:17

Design and Implementation of Matrix Transposition Unit for Real-Time SAR Image Systems
下载PDF
导出
摘要 矩阵转置是多维图像和信号处理中常用的处理过程 ,在SAR成像系统中也有非常重要的应用 重点讨论了数据流处理中行进列出的一类转置问题 ,结合SAR成像算法 ,分析成像系统中矩阵转置的特点 ,提出了输入输出平衡的数据流转置方式 ,解决由于硬件实现中存储器读写速度不同和输入输出数据量不同所带来的输入输出负载不平衡问题 还介绍了SAR成像系统中转置存储部件的实现问题 转置存储部件采用双总线的结构 ,图像数据的输入和输出可以并行进行 ,具有非常高的数据传输速度 ,在 6 6MHz的工作频率下 ,能达到 Matrix transposition is a fundamental operation in multidimensional image and signal processing, especially in SAR image systems With emphasis on the row-in and column-out matrix transposition problem in data flow prcessing and based on the SAR image algorithm, the matrix trasposition problem is invesigated in detail, and a balance matrix transposition method for data input and output is proposed This method can solve the imbalancing load problem due to unequal memory access speed and unequal data amount of input and output A high-speed matrix transposition module is introduced, which uses 2 data bus:input bus and output bus Data input and output can be processed parallelly The results indicate that it has the peak performance of 340MB/s in 66?MHz clock
出处 《计算机研究与发展》 EI CSCD 北大核心 2003年第1期6-11,共6页 Journal of Computer Research and Development
基金 国家自然科学基金重大项目 (698962 5 0 ) 中国科学院计算技术研究所青年基金 (2 0 0 162 80 2 )
关键词 SAR 成像系统 矩阵转置 设计 合成孔径雷达 数据流处理 信号处理 SAR data flow processing matrix transposition
  • 相关文献

参考文献9

  • 1黄煦涛主编, 胡光锐等译. 二维数字信号处理Ⅱ变换与中值滤波器. 北京:科学出版社,1985(Thomas S Huang. Two-Dimensicnal Digital Signal Processing Ⅱ, Berlin: Springer, 1981)
  • 2S Panchanathan. Universal architecture for matrix transposition. IEE Proceedings-E, 1992, 139(5): 387~392
  • 3J A Tierno, P Kudva. Asynchronous transpose-matrix architectures. The Int'l Conf on Computer Design, Austin, TX, 1997
  • 4谢应科 史云涛 等.实时高分辨率SAR成像系统中CTM的设计[J].计算机研究与发展,2000,37:195-199.
  • 5S Chatterjee, S Sen. Cache-efficient matrix transposition. High-Performance Computer Architecture, 2000, HPCA-6: 195~205
  • 6Francescomaria Marino, Earl E Swartzlander, Jr. Parallel implementation of multidimensional transforms without interprocessor communication. IEEE Trans on Computers, 1999, 48(9): 951~961
  • 7Jinwoo Suh, Viktor K Prasanna. An efficient algorithm for large-scale matrix transposition. The 2000 Int'l Conf on Parallel Processing, Toronto, 2000
  • 8M Kandemir, A Choudhary, J Ramanujam et al. A matrix-based approach to the global locality optimization problem. The Int'l Conf on Parallel Architectures and Compilation Techniques, Paris, France, 1998
  • 9Kyeounsoo Kim, Peter A Beerel. A low-power matrix transposer using MSB-controlled inversion coding. IEEE Asia-Pacific ASIC Conf, Seoul, Korea, 1999

共引文献3

同被引文献82

引证文献17

二级引证文献41

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部