国防科技大学学报2017，Vol.39Issue(5)：50-55,6.DOI:10.11887/j.cn.201705009
采用向量内积的并行相关算法
Parallel correlation algorithm based on vector dot product
摘要
Abstract
For solving the real-time problem of the software correlator,a parallel receiving model of spread spectrum signal based on vector dot product was proposed.The calculation speed of the signal correlation was improved by parallel computing correlation value of each channel with a large number of ALU in GPU (graphics processing unit).Simulation results show that,25 channels with 150 correlation computation consume 967 μs in all for 1 ms signal of 25 MHz sampling rate by using the vector dot product parallel correlation algorithms based on GPU.Compared with the implementation on CPU based on M KL(math kernel library),the proposed method speeds up to 61.4 times,and the real-time correlation of the broadband spread spectrum signals in software can be realized.关键词
相关器/向量内积/软件接收机/并行计算/图形处理单元Key words
correlator/vector dot product/software receiver/parallel computing/GPU分类
信息技术与安全科学引用本文复制引用
牟卫华,倪少杰,白洋,孙广富,欧钢..采用向量内积的并行相关算法[J].国防科技大学学报,2017,39(5):50-55,6.基金项目
国家自然科学基金资助项目(41604016) （41604016）
