Optimization of Clutter Simulation Based on GPU

Guowei Xu,Fangrui Hao,Jian Chen,Chunbo Xiu

doi:10.1109/access.2020.2972941

Guowei Xu, Fangrui Hao + Show 2 more

Open Access

https://doi.org/10.1109/access.2020.2972941

Copy DOI

Abstract

Weibull clutter is used as an example in this paper. Based on the serial parallel analysis of Zero-memory non-linear transformation's Weibull distributed clutter algorithm, fine-grained optimization is performed. The fine-grained part uses the cuBLAS library to optimize the performance of convolution calculations. Compared with CUDA shared memory convolution method and GPU parallel matrix multiplication convolution method, its computational performance can be significantly improved under a large amount of data. Simulation results show that the Zero-memory non-linear transformation's Weibull distributed clutter simulation method is optimized and accelerated. The real-time performance of clutter data is significantly improved and its acceleration effect will be better as the amount of clutter data increases. It turns out that through fine-grained optimization, the performance of convolution calculations with large amounts of data is improved.

Highlights

With the development of GPU and CUDA parallel computing, when faced with a large amount of data processing, GPUs increase the processing speed which has promoted the development of many fields
How to construct a ground clutter model that is highly similar to the actual environment is one of the important problems faced at this stage [6]
EXPERIMENTAL RESULTS AND ANALYSIS The sampling frequency is set to Fs = 1000Hz, f3dB = 50Hz, the shape of the power spectrum is Gaussian power spectrum, and the sampling points are set to 1 × 105

Summary

INTRODUCTION

With the development of GPU and CUDA parallel computing, when faced with a large amount of data processing, GPUs increase the processing speed which has promoted the development of many fields. Such as medical data processing [1], radar signal processing [2] and deep learning. The fine-grained optimization of convolution under large amounts of data is mainly implemented to greatly improve the performance of convolution calculation. The Weibull distribution clutter simulation acceleration of ZMNL is realized which improves the real-time performance of clutter data simulation.

CLUTTER SIMULATION PRINCIPLE

EXPERIMENTAL RESULTS AND ANALYSIS

CONCLUSION