基于DSP的被动声纳波束形成系统设计实现
发布时间：2018-10-04 22:21
【摘要】：波束形成技术是声纳信号处理中重要的组成部分,其性能直接影响信号处理系统的整体指标。波束形成技术具有运算量大以及实时处理等特点,因此研发高速度高性能的波束形成器很有必要。本文从工程实现的角度出发,设计并实现了一种基于TMS320C6678多核DSP的被动声纳波束形成系统。本文首先介绍了被动声纳和波束形成的背景知识,分析时域波束形成和频域波束形成原理,探讨了一种基于线性调频Z变换多波束形成技术,并对算法进行了 MATLAB仿真和验证。根据项目的技术指标,首先给出被动声纳波束形成处理的整体方案。研究了 TMS320C6678的PCIe模块实现了与上位机的通信。为了满足实时性要求,给出了一种基于多核DSP实现CZT多波束形成的解决方案。在数据处理过程中使用了 EDMA3完成数据转置和数据的搬移,将大尺寸一维FFT转换为小尺寸二维FFT计算,大大缩短了数据处理时间,满足了系统实时性要求。此外,本文还对核间通信、Cache一致性问题和多核DSP的启动加载进行了研究。最后对TMS320C6678上的各个模块进行了调试,并将数据处理结果与MATLAB处理结果进行对比,验证了数据的正确性。
[Abstract]:Beamforming technology is an important part of sonar signal processing, and its performance directly affects the whole index of signal processing system. Beamforming technology has the characteristics of large computation and real-time processing, so it is necessary to develop high speed and high performance beamformer. In this paper, a passive sonar beamforming system based on TMS320C6678 multi-core DSP is designed and implemented from the point of view of engineering implementation. This paper first introduces the background knowledge of passive sonar and beamforming, analyzes the principle of time-domain beamforming and frequency-domain beamforming, discusses a multi-beamforming technique based on linear frequency modulation Z transform, and simulates and verifies the algorithm by MATLAB. According to the technical specifications of the project, the overall scheme of passive sonar beamforming processing is first presented. The communication between the PCIe module of TMS320C6678 and the host computer is studied. In order to meet the real-time requirement, a solution of CZT multi-beam forming based on multi-core DSP is presented. In the process of data processing, EDMA3 is used to complete the data transposition and data transfer, and the large-size one-dimensional FFT is converted into the small-size two-dimensional FFT calculation, which greatly shortens the data processing time and meets the real-time requirements of the system. In addition, the consistency problem of inter-core communication cache and the startup loading of multi-core DSP are also studied in this paper. Finally, every module on TMS320C6678 is debugged, and the result of data processing is compared with that of MATLAB to verify the correctness of the data.
【学位授予单位】：南京理工大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TB56
[Abstract]:Beamforming technology is an important part of sonar signal processing, and its performance directly affects the whole index of signal processing system. Beamforming technology has the characteristics of large computation and real-time processing, so it is necessary to develop high speed and high performance beamformer. In this paper, a passive sonar beamforming system based on TMS320C6678 multi-core DSP is designed and implemented from the point of view of engineering implementation. This paper first introduces the background knowledge of passive sonar and beamforming, analyzes the principle of time-domain beamforming and frequency-domain beamforming, discusses a multi-beamforming technique based on linear frequency modulation Z transform, and simulates and verifies the algorithm by MATLAB. According to the technical specifications of the project, the overall scheme of passive sonar beamforming processing is first presented. The communication between the PCIe module of TMS320C6678 and the host computer is studied. In order to meet the real-time requirement, a solution of CZT multi-beam forming based on multi-core DSP is presented. In the process of data processing, EDMA3 is used to complete the data transposition and data transfer, and the large-size one-dimensional FFT is converted into the small-size two-dimensional FFT calculation, which greatly shortens the data processing time and meets the real-time requirements of the system. In addition, the consistency problem of inter-core communication cache and the startup loading of multi-core DSP are also studied in this paper. Finally, every module on TMS320C6678 is debugged, and the result of data processing is compared with that of MATLAB to verify the correctness of the data.
【学位授予单位】：南京理工大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TB56
【相似文献】
相关期刊论文 前10条
1 李峰;王毅;李枚毅;;基于小波的被动声纳降噪研究[J];东华理工学院学报;2007年04期
2 王丹;闫利超;;被动声纳信号检测技术研究[J];信息安全与通信保密;2009年09期
3 蒋小勇;于李洋;杜选民;;提高被动声纳宽带检测性能的一种新方法[J];声学技术;2011年06期
4 储招江;数字计算式被动声纳系统[J];声学与电子工程;1989年04期
5 黄凡;韩树平;;基于方位与时延信息融合的被动声纳航迹关联[J];计算机仿真;2008年05期
6 王学志;吴克桐;Q糜，
本文编号：2252100
本文链接：https://www.wllwen.com/guanlilunwen/gongchengguanli/2252100.html
