摘要
首先,针对噪声不存在有限方差的问题,对标准归一化相关函数的定义域进行了扩展,实现了背景噪声的抑制。在此基础上,利用谐波信号频率与噪声子空间的正交性,提出了谐波信号频率估计的MUSIC方法。仿真结果表明,与分数低阶类方法相比,由于标准归一化相关函数是一个线性算子,极大地简化了稳定分布信号处理的计算和分析过程。
A standard normalized correlation-based MUSIC method is developed for harmonic signal frequency estimation inαnoise.First,as there is no finite variance in theαnoise,the domain of the standard normalized correlation function is extended,and theαnoise is suppressed by this function.Based on these,the MUSIC method for harmonic signal frequency estimation is proposed using the orthogonality between the harmonic signal frequency and the noise subspace.Compared with the fractional lower order classification method,the standard normalized correlation function is a linear operator,which greatly simplifies the calculation and analysis of theα stable distributed signal processing.
出处
《吉林大学学报（工学版）》
EI
CAS
CSCD
北大核心
2016年第3期923-928,共6页
Journal of Jilin University:Engineering and Technology Edition
基金
国家自然科学基金项目(61171079)
关键词
信号处理
稳定分布
谐波恢复
归一化相关函数
signal processing
stable distribution
harmonic recovery
standard normalized correlation function
