频率估计的一种同频信号加权融合算法  

Weight-fusion Algorithm for Frequency Estimation of Co-frequency Signal

作　　者：肖玮[1] 涂亚庆[1] 刘良兵[1] 闾琳[2] 

机构地区：[1]后勤工程学院后勤信息工程系,重庆401311 [2]成都电子机械高等专科学校,成都611730

出　　处：《系统仿真学报》2013年第4期742-747,752,共7页Journal of System Simulation

基　　金：国家自然科学基金(61271449;61201450;60871098);重庆市自然科学基金(cstc2012jjA0877;cstc2012jjA1047;cstc2011BA2015)

摘　　要：为提高低信噪比条件下短时正弦信号的频率估计精度,针对任意长度的正弦信号,提出一种同频信号加权融合算法。首先,给出同频信号的定义和频谱模型;其次,构造具有相位连续化特性和噪声对消特性的相位修正矩阵对同频信号频谱进行加权融合,使加权融合后的频谱近似于相位连续信号频谱。最后,谱峰搜索加权融合后的频谱,获得高精度的频率估计值。仿真实验表明,与现有方法相比,算法抗噪性强,普适性好,频率估计精度高。In order to improve the frequency estimation of any length sinusoid signal at low signal-to- noise ratio （SNR）, a frequency estimation algorithm based on weight-fusion of the co-frequency signal was proposed. Firstly, the definition of the co-frequency signal and its spectrum model were given. Secondly, the phase compensating matrix was constructed, which could make phases coherent and noise reduction. And the spectrum of the co-frequency signal was weight-fused by this proposed matrix to turn almost the same as the spectrum of the phase-coherent signal. Consequently, high frequency estimation precision could be obtained through spectral peak searching of the weight-fusion spectrum. Algorithm simulation results show that this proposed algorithm has strong noise immunity, good universality and high precision of frequency estimation compared with the present methods.

关 键 词：频率估计 加权融合 同频信号 噪声对消 低信噪比 算法仿真 

分 类 号：TN957.51[电子电信—信号与信息处理]