基于自适应波束形成和ICA的消噪系统

Noise cancellation system based on adaptive beamforming and independent component analysis

介绍了自适应宽带波束形成和快速独立分量分析的基本理论和算法,分析了固定波束形成的理论局限性及自适应宽带波束形成的优越性。提出了一种基于自适应波束形成和独立分量分析的消噪系统,算法显著地抑制了噪声、增强了语音,又具有稳定快速的性能。同时分析了自适应波束形成单元数或者输入信号数对算法性能的影响,对实际应用具有指导意义。

This paper presents a new hybrid noise cancellation scheme employing Adaptive Beamforming and Fast Independent Component Analysis. The method can restrain noise and enhances speech with stability and fast convergence. Furthermore the influence upon this algorithm when the number of units or input signals of adaptive beamforming varies is studied, which is significant to practical application .