摘要
针对强噪声背景下线性调频混叠信号,提出了一种结合分数阶傅里叶变换的盲源分离算法,能够有效提升强噪声背景下的信号提取性能。首先将混叠信号转换到分数阶傅里叶域并估计噪声功率谱,然后在最优阶域下对混叠信号使用谱减法,最后采用快速独立分量分析算法对混叠信号进行盲分离并平滑滤波。仿真结果表明所提算法有效提升了强噪声背景下线性调频信号的盲分离性能,对比2种算法的分离波形以及信号均方误差,该算法的分离效果较独立分量分析算法有明显改善。
Aiming at linear frequency modulation （LFM） mixed signals in the background of strong noise,this paper puts forward a blind source separation algorithm combined with fractional Fourier transform,which can effectively improve the signal extraction performance under the strong noise background. Firstly, the mixed signal is transformed into the fractional Fourier domain and the noise power spectrum is estimated. Then, the spectral subtraction is used for the mixed signal under the optimal order domain. Finally, the fast independent component analysis algorithm is used to perform the blind separation and smooth filtering for the mixed signals. Simulation results show that the proposed algorithm can effectively improve the blind separation performance of LFM signal in the background of strong noise. The separation waveform and signal mean square error of two algorithms are compared, and the separation effect of proposed algorithm is improved obviously compared with the independent component analysis algorithm.
出处
《舰船电子对抗》
2017年第6期75-79,90,共6页
Shipboard Electronic Countermeasure
基金
国家自然科学基金
项目编号:61601491
关键词
线性调频信号
分数阶傅里叶变换
最优阶
谱减法
快速独立分量分析
linear frequency modulation signal
fractional Fourier transform
optimal order
subtraction of spectrum
fast independent component analysis
