XAO OpenIR  > 计算机技术应用研究室
基于零拷贝的脉冲星GPU相干消色散算法
Alternative TitleCoherent De-dispersion Algorithm for Pulsar GPU Based on Zero-copy
王博群1,2; 张海龙1,2,3,4; 王杰1,2,4; 冶鑫晨1,4; 王万琼1; 李嘉1; 张萌1,2; 张亚州1,2
2021-07-01
Source Publication天文研究与技术
ISSN1672-7673
Volume18Issue:3Pages:372-379
Contribution Rank1
Abstract射电脉冲信号在传输过程中受到星际介质的影响会导致轮廓展宽和变形,在研究过程中需要对信号进行消色散处理。本文设计并实现了基于零拷贝的脉冲星数据GPU相干消色散算法,采用设备内存映射以消除主机到设备的拷贝开销,利用CUDA的cuFFT库进行多BATCH傅里叶变换以提高DFT效率,同时采用多线程实现了传递函数的加速计算。实验结果表明,与传统CPU及GPU算法相比,本文提出的算法在大数据量时表现良好。
Other AbstractIn order to meet the requirements of real-time de-dispersion processing of pulsar signals, a GPU based incoherent de-dispersion algorithm is implemented. We use high-performance parallel computing methods to conduct in-depth research on the incoherent de-dispersion algorithm for multi-thread processing, and propose a parallel acceleration scheme for the algorithm, which solves the problem that the de-dispersion algorithm has a large amount of calculation and cannot be processed in real time. We analyze the intensive computing parts of the algorithm, efficiently use the hierarchical storage structure of the GPU, improve the utilization rate of GPU computing resources, thereby reducing the calculation time, and significantly improving the calculation performance of the incoherent de-dispersion algorithm.
Keyword零拷贝 相干消色散 GPU CUDA cuFFT
DOI10.14005/j.cnki.issn1672-7673.20210107.001
Indexed ByCSCD
Language中文
WOS IDCSCD:7008058
CSCD IDCSCD:7008058
Citation statistics
Document Type期刊论文
Identifierhttp://ir.xao.ac.cn/handle/45760611-7/4687
Collection计算机技术应用研究室
Corresponding Author张海龙
Affiliation1.中国科学院新疆天文台,新疆乌鲁木齐830011;
2.中国科学院大学,北京100049;
3.中国科学院射电天文重点实验室,江苏南京210008;
4.国家天文科学数据中心,北京100101
First Author AffilicationXinjiang Astronomical Observatory, Chinese Academy of Sciences
Corresponding Author AffilicationXinjiang Astronomical Observatory, Chinese Academy of Sciences
Recommended Citation
GB/T 7714
王博群,张海龙,王杰,等. 基于零拷贝的脉冲星GPU相干消色散算法[J]. 天文研究与技术,2021,18(3):372-379.
APA 王博群.,张海龙.,王杰.,冶鑫晨.,王万琼.,...&张亚州.(2021).基于零拷贝的脉冲星GPU相干消色散算法.天文研究与技术,18(3),372-379.
MLA 王博群,et al."基于零拷贝的脉冲星GPU相干消色散算法".天文研究与技术 18.3(2021):372-379.
Files in This Item:
File Name/Size DocType Version Access License
王博群-2021-基于零拷贝的脉冲星GP(1543KB)期刊论文出版稿开放获取CC BY-NC-SAView Application Full Text
Related Services
Recommend this item
Bookmark
Usage statistics
Export to Endnote
Google Scholar
Similar articles in Google Scholar
[王博群]'s Articles
[张海龙]'s Articles
[王杰]'s Articles
Baidu academic
Similar articles in Baidu academic
[王博群]'s Articles
[张海龙]'s Articles
[王杰]'s Articles
Bing Scholar
Similar articles in Bing Scholar
[王博群]'s Articles
[张海龙]'s Articles
[王杰]'s Articles
Terms of Use
No data!
Social Bookmark/Share
File name: 王博群-2021-基于零拷贝的脉冲星GPU相干消色散算法.pdf
Format: Adobe PDF
This file does not support browsing at this time
All comments (0)
No comment.
 

Items in the repository are protected by copyright, with all rights reserved, unless otherwise indicated.