高级检索

噪声环境下水声信号瞬时盲源分离方法研究

Instantaneous blind source separation of underwater acoustics signals in a noisy background

  • 摘要: 噪声环境下大多数盲源分离算法性能大大降低。提出了一种适用于噪声环境的基于子带分解的瞬时盲源分离算法。通过修正白化矩阵和选取合理子带降低盲源分离过程中由噪声产生的不利影响。仿真实验表明,采用基于子带分解的瞬时盲源分离算法能够较好地实现噪声环境下水声混合信号的分离;进一步研究表明,该算法在噪声环境下表现出良好的稳健性。

     

    Abstract: Most algorithms for blind signal separation(BSS) have poor performance in a noisy back-ground.An instantaneous blind signal separation method based on subband decomposition is presented to deal with noisy BSS.Using a modified whitening matrix and selecting an appropriate subband,draw-backs due to background noise can be alleviated.Simulation experiments show that mixed underwater acoustic signals in noisy background can well be separated.In addition,the method is robust in the back-ground noise of different intensities.

     

/

返回文章
返回