期刊文献+

基于C6678多核数字信号处理器的声纳信号并行处理设计 被引量:5

Sonar Signal Processing Parallel Design Based on C6678 Multicore DSP
下载PDF
导出
摘要 针对多核并行机制下,共享资源争夺激烈,硬件能力提升难以切实转变成程序效率提高的难题,通过协调存储器访存和核间同步等关键问题,研究了一种基于C6678多核数字信号处理器的声纳信号多级并行处理方法,包括核间流水线设计、数据传输与中央处理器并行设计和指令流线设计。以声纳二维相控方位滤波为例,介绍各级设计的实现方法,逐个测试并行性能,并编制实时处理软件。测试结果表明,该方法能够实现存储器访问和中央处理器运算并行,极大提高程序执行效率。通过采用该方法开发的实时处理系统,具有集成度高和实时性强的优点,获得了高航速下海中浮球和配重条石清晰的实时成像效果,具有工程应用价值。 In parallel processing system with multicore,it is hard to turn the improvement of hardware into the efficiency of software substantially for the drastic competition of shared resources on chip. To solve the problem,a multistage parallel sonar signal processing method based on C6678 multicore digital signal processor( DSP) is presented by scheduling memory access and multicore synchronization,which includes multicore pipeline design,parallel processing design of data transmission and CPU computation,and assembly code pipeline design. The multi-parallel stage method is introduced by taking two-dimension phased azimuth filter as an example. Performance of each stage above is tested,and a real-time processing software is developed. Test result shows that the proposed method can parallel memory access and CPU computation effectively,and the software developed based on the proposed method works efficiently.In a high speed sea trial,the sharp images of a suspension ball and its counterweight rectangle stone were got by the real-time processing system. The multistage parallel processing method is of great application value in engineering.
出处 《兵工学报》 EI CAS CSCD 北大核心 2016年第8期1476-1481,共6页 Acta Armamentarii
基金 武器装备预先研究项目(7301303)
关键词 声学 声纳 相控方位滤波 多核数字信号处理器 并行处理 acoustics sonar phased azimuth filter multicore DSP parallel processing
  • 相关文献

参考文献8

二级参考文献46

  • 1Hill M D, Marry M R. Amdahl's law in the multicore era[J]. Computer, 2008,41 (7) : 33 -- 38.
  • 2Michaud R P. Replacement policies for shared caches on symmetric multicores:a programmer-centrie point of view[J ].IRISA, 2008,21( 1 ) :3 - 19.
  • 3Dybdahl H, Natvig L. A cache-partitioning aware replacement policy for chip multiprocessors[C]//Proceeding of 2006 ACM Conference on High Performance Computing. Bangalore: Springer, 2006:22 -34.
  • 4Qureshi M K, Patt Y N. Utility-based cache partitioning: a low-overhead, high performance runtime mechanism to partition shared cache[C]//Proceeding of the 39th Annual IEEE/ACM International Symposium on Microarchitecture. Orlando: IEEE, 2006:423 -432.
  • 5Zhao L, lyer R, Upton M, et al. Hybrid last level caches for chip-multiprocessors [ J ]. ACM SIGARCH Computer Architecture, 2008,36 (2) : 56 - 63.
  • 6Lai S C, Lu S L, Stark J, et al. Cache misses for accurate data speculation and prefetching [ C ] // Proceeding of International Conference on Supercomputing. New York: ACM, 2002:189- 198.
  • 7Standard Performance Evaluation Corporation. SPEC CPU 2000[EB/OL]. [2007- 02-01]. http://www, spec. org.
  • 8Hsu R, Reinhardt K, Iyer R, et al. Communist, utilitarian, and capitalist cache policies on CMPs: Caches as a shared resource [C] //Proc of the 15th Int Conf on Parallel Architecture and Compilation Techniques (PACT'06). New York: ACM, 2006:13-22.
  • 9Suh G, Srinivas D, Rudolph L. A new memory monitoring scheme for memory aware scheduling and partitioning [C] //Proc of the 8th Int Syrup on High Performance Computer Architecture ( HPCA'02 ). Piscataway, NJ: IEEE, 2002:117-128.
  • 10Suh G, Devadas S, Rudolph L. Dynamic partitioning of shared cache memory[J]. The Journal of Supercomputing, 2004, 28(1) : 7-26.

共引文献59

同被引文献38

引证文献5

二级引证文献7

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