基于零拷贝的脉冲星GPU相干消色散算法  

Coherent De-dispersion Algorithm for Pulsar GPU Based on Zero-copy

作　　者：王博群 张海龙[1,2,3,4] 王杰 冶鑫晨[1,4] 王万琼 李嘉 张萌[1,2] 张亚州[1,2] Wang Boqun;Zhang Hailong;Wang Jie;Ye Xinchen;Wang Wanqiong;Li Jia;Zhang Meng;Zhang Yazhou(Xinjiang Astronomical Observatory,Chinese Academy of Sciences,Urumqi 830011,China;University of Chinese Academy of Sciences,Beijing 100049,China;Key Laboratory of Radio Astronomy,Chinese Academy of Sciences,Nanjing 210008,China;National Astronomical Data Center,Beijing 100101,China)

机构地区：[1]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

出　　处：《天文研究与技术》2021年第3期372-379,共8页Astronomical Research & Technology

基　　金：国家自然科学基金(11873082,11803080);国家重点研发计划(2018YFA0404704);中国科学院青年创新促进会;国家天文科学数据中心;中国科学院科学数据中心体系资助.

摘　　要：射电脉冲信号在传输过程中受到星际介质的影响,脉冲轮廓展宽变形,在研究过程中需要对信号进行消色散处理。设计并实现了基于零拷贝的脉冲星数据图形处理器(Graphic Processing Unit,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,利用统一计算设备架构(Compute Unified Device Architecture,CUDA)的cuFFT库进行多批次傅里叶变换以提高离散傅里叶变换(Discrete Fourier Transform,DFT)的效率,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,与传统中央处理器及图形处理器算法相比,本文的算法在数据量大时表现良好。The radio pulse signal is affected by the interstellar medium during the transmission process,which will cause the profile to be broadened and deformed.In the research process,the signal needs to be de-dispersed.This paper designs and implements a zero-copy-based GPU coherent de-dispersion algorithm for pulsar data,uses device memory mapping to eliminate the copy cost from host to device,and uses CUDA′s cuFFT library to perform multi-BATCH Fourier transform to improve DFT efficiency.Multithreading is used to realize the accelerated calculation of the transfer function.The experimental results show that,compared with traditional CPU and GPU algorithms,the algorithm proposed in this paper performs well in large amounts of data.

关 键 词：零拷贝 相干消色散 图形处理器 统一计算设备架构 cuFFT 

分 类 号：P161.4[天文地球—天文学]