基于GPU的MTD性能优化

为了解决传统雷达信号处理机在研发阶段面临的调试困难,计算能力受硬件限制及程序复用性差等问题,本文提出了使用GPU作为雷达计算核心的方案.在使用GPU实现雷达信号处理算法的过程中,动目标检测(MTD)部分的优化效果远低于脉冲压缩和恒虚警检测.经过分析,MTD过程中的矩阵转置与向量点乘占据了算法的大量时间.本文从GPU的数...

Performance Optimization of MTD Based on GPU

In order to solve the problems of adjustment,hardware limitation and program reusability of traditional radar signal processor in the research and development stage,this paper proposes a scheme of using GPU as the computing core of radar.In the process of using GPU to realize radar signal processing algorithms,the improvement in moving target detection(MTD)is far less than that in PC and CFAR.Our analysis reveals that the matrix transpose and vector point multiplication occupy a lot of time in the MTD process.This paper starts with the data reading mode of GPU and the characteristics of CUDA function,optimizes the process of fast Fourier transform based MTD,and uses CUBLAS matrix operation on GPU to enable data filtering after pulse compression by finite impulse response filter bank,so as to realize more flexible MTD.Compared with the results of a CPU platform,the error of the final GPU calculation results is less than 0.05%,and the performance improvement is 200 times of that of the CPU platform.