欢迎访问《兵工学报》官方网站,今天是 分享到:

兵工学报 ›› 2019, Vol. 40 ›› Issue (8): 1725-1731.doi: 10.3969/j.issn.1000-1093.2019.08.023

• 论文 • 上一篇    下一篇

基于分布式压缩感知的麦克风阵列声源定位

黄惠祥, 郭秋涵, 童峰   

  1. (厦门大学 水声通信与海洋信息技术教育部重点实验室, 福建 厦门 361100)
  • 收稿日期:2018-10-16 修回日期:2018-10-16 上线日期:2019-10-15
  • 通讯作者: 童峰(1973—),男,教授,博士生导师 E-mail:ftong@xmu.edu.cn
  • 作者简介:黄惠祥(1993—),男,硕士研究生。E-mail:497242232@qq.com
  • 基金资助:
    福建省高校产学合作项目(2015H6019);中央高校基本科研业务费专项资金项目(20720190102)

Microphone Array Sound Source Direction-of-arrival Estimation Based on Distributed Compressed Sensing

HUANG Huixiang, GUO Qiuhan, TONG Feng   

  1. (Key Laboratory of Underwater Acoustic Communication and Marine Information Technology of Ministry of Education, Xiamen University, Xiamen 361100, Fujian, China)
  • Received:2018-10-16 Revised:2018-10-16 Online:2019-10-15

摘要: 压缩感知-正交匹配追踪(CS-OMP)算法将声源定位问题转化为信号稀疏重构问题,能比传统定位算法获得更高的定位性能。但是CS-OMP算法在定位中没有考虑多个信号的相关性。将分布式压缩感知(DCS)理论引入麦克风阵列的声源定位中,考虑信号之间具有的共同稀疏性,利用分布式压缩感知-同步正交匹配追踪(DCS-SOMP)算法对信号进行联合重构,获取稀疏位置并对声源实现定位。仿真和实验结果表明,与传统定位算法和CS-OMP算法相比,DCS-SOMP算法在低信噪比环境下具有更好的定位性能和鲁棒性。

关键词: 麦克风阵列, 声源定位, 分布式压缩感知, 稀疏信号, 联合重构

Abstract: Compressed sensing-orthogonal matching pursuit (CS-OMP) algorithm is to transform the sound source localization problem into a signal sparse reconstruction problem, which can obtain higher positioning performance than the traditional positioning algorithm. However, the CS-OMP algorithm does not consider the correlation of multiple signals in positioning. The distributed compressed sensing (DCS) theory is introduced into the sound source localization of microphone array. Considering the common sparsity between the signals, the distributed compressed sensing-simultaneous orthogonal matching pursuit (DCS-SOMP) algorithm is used to reconstructure the signals for constructing a sparse position and positioning the sound source. Simulated and experimental results show that the DCS-SOMP algorithm has better positioning performance and robustness in low SNR environment compared with the traditional positioning algorithm and CS-OMP algorithm. Key

Key words: microphonearray, soundsourcelocalization, distributedcompressedsensing, sparsesignal, jointreconstruction

中图分类号: