期刊文献+

水声信号盲分离的实数算法与复数算法 被引量:8

Easily Implementable Complex Domain Algorithms of BlindSource Separation for Underwater Acoustic Array
在线阅读 下载PDF
导出
摘要 分析了实数算法和复数算法实现水声信号盲分离的原理,对几种没有复数形式的盲分离算法极出了修改,修改后的算法可以完成复数信号的盲分离。在三元均匀线列阵上对几种典型的算法如 FASTICA、ACY和 EASI算法分离水声信号的性能进行了仿真。结果表明,在精确补偿延时形成波束的条件下实数算法性能与复数算法一致;在基阵几何尺寸存在误差和延时补偿不精确的条件下,实数算法无效,复数算法仍保持良好的性能。 Abstract: To our best knowledge, only EASI complex domain algorithm is easily implementable up to now. We now propose other easily implementable complex domain algorithms of blind source separation for underwater acoustic array. We analyze the principles of blind separation of underwater acoustic signals with real domain and complex domain algorithms. We modify several real domain algorithms to make them easily implementable complex domain ones. Through computer simulation, in which a three sensor linear array with uniform spacing is considered, we study the performance of .typical algorithms such as ACY (real domain) and EASI (complex domain). Geometrical error of the array and time-delay error of compensation inevitably exist and they cause real domain algorithms fail to converge as shown by upper curves of Figs. 2(a) and 2(b) for EASI and ACY respectively. The ACY complex domain algorithm we propose can still converge even when geometrical error of array and time-delay error of compensation exist, as shown by lower curve of Fig. 2(b). We have modified real domain MHJ and BG algorithms to make them easily implementable complex domain ones, and we believe they can also converge when geometrical and time-delay errors inevitably exist.
出处 《西北工业大学学报》 EI CAS CSCD 北大核心 2002年第1期44-48,共5页 Journal of Northwestern Polytechnical University
基金 国家科学自然基金(60072052)资助
关键词 盲源分离 均匀线列阵 水声信号 实数算法 复数算法 信号处理 Key words: blind source separation(BSS), complex domain algorithm, underwater acoustic array
  • 相关文献

参考文献8

  • 1[1]Li S, Sejnowski T J. Adaptive Separation of Mixed Broadband Sound Sources with Delays by a Beamforming Herault-Jutten Network. IEEE J of Oceanic Engineering, 1995, 20(1): 73~79
  • 2[2]Cardoso J F, Laheld B. Equivariant Adaptive Source Separation. IEEE Transaction on Signal Processing,1996,44(12): 3017~3030
  • 3[3]Amari S-I, Cichocki A, Yang H H. A New Learning Algorithm for Blind Signals Separation. In: Touretzky D, Mozer, Hasselmo M. Advances in Neural Information Processing Systems. Massachusetts: MIT Press, 1996, 757~763
  • 4[4]Hyvarinen A, Oja E. A Fast Fixed-Point Algorithm for Independent Component Analysis. Neural Computation, 1997, 9(7): 1483~1492
  • 5[5]Bingham E, Hyvarinen A. A Fast Fixed-Point Algorithm for Independent Component Analysis of Complex Valued Signals. Int J of Neural Systems, 2000, 10(1): 1~8
  • 6[6]Karhunen J, Oja E, Wang L, Vigaro R, Joutsensalo J. A Class of Neural Networks for Independent Component Analysis. IEEE Trans on Neural Networks, 1997, 8(3): 486~504
  • 7[7]Cichocki A, Bonger R E, Moszczynski L, Pope K. Modified Herault-Jutten Algorithms for Blind Separation of Sources. Digital Signal Processing, 1997, 7(1): 80~93
  • 8[8]Comon P. Independent Component Analysis, a New Concept? Signal Processing, 1994, 36(2): 287~314

同被引文献98

引证文献8

二级引证文献30

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部