摘要
为提高在低信噪比和存在小信号情况下的源信号数目估计正确率,提出了一种利用ICA估计出的信号和利用空间谱分离出的信号的相关系数最大化为依据的源数目估计方法。该方法综合利用源信号的统计独立特性和阵列结构信息,降低了算法对噪声和信号强弱的敏感。计算机仿真结果表明,在低信噪比和存在弱信号的情况下,该算法可以高精度的估计出源信号数目。
To improve the source number estimation accuracy for low SNR and small-signal case, a method for source number estimation, that use the maximize correlation coefficient of signal waveforms were estimated by ICA and signal waveforms were separated by spatial Spectrum is proposed. The array of structural information and statis- tical independence properties of the source signal is used to the proposed method, thus reducing the algorithm sen- sitive to noise and signal strength. Computer simulation results show that the algorithm can be highly accurate esti- mates of the number of source signals in the low SNR and small-signal case.
出处
《科学技术与工程》
北大核心
2012年第35期9507-9512,共6页
Science Technology and Engineering
