首页> 中文期刊> 《计算机工程》 >基于SIMD架构的相干累加运算优化方法

基于SIMD架构的相干累加运算优化方法

         

摘要

A new method to optimize the computing program of coherent integration based on Single Instruction Multiple Data(SIMD) is proposed to deal with the problem of real-time in the software GPS receiver, and the method of SIMD and pipeline optimization is discussed respectively. This paper analyzes the structure of following loop and SIMD instructs and coherent integration accumulation operation quantity, and uses Multi Media eXtension(MMX) order set to optimize the pipeline. Experimental results show that, the time of completing process of 1 second length data is about 0.7 second. The computation is only 6.5% of general Single Instruction Single Data(SISD) instruction, which effectively ensures the implementation of pure real-time software receiver.%针对软件GPS接收机在处理高数据流时存在的实时性问题,提出一种基于单指令多数据流(SIMD)架构的相干运算优化方法.分.析跟踪环路结构、SIMD指令与相干积分累加运算量,使用多媒体扩展指令集对流水线进行优化.实验结果表明,完成1 s数据的相干积分累加计算时间为0.7 s,该运算量仅为使用普通单指令多数据流指令的6.5%,具有较高的实时性.

著录项

相似文献

  • 中文文献
  • 外文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号