低信噪比下基于自相关函数的频率估计方法  

Frequency estimator based on autocorrelation with low SNR

作　　者：周磊磊[1] 罗炬锋[1] 付耀先[1] 王营冠[1] 

机构地区：[1]中国科学院上海微系统与信息技术研究所无线传感器网络与通信重点实验室,上海201800

出　　处：《华中科技大学学报（自然科学版）》2014年第4期45-49,共5页Journal of Huazhong University of Science and Technology(Natural Science Edition)

基　　金：上海市科委技术标准专项资助项目(12DZ0500100)

摘　　要：提出一种基于自相关函数的频率估计方法.对归一化的自相关函数计算相邻共轭乘积,通过近似处理将乘积表达式中信号与噪声的相乘关系转化为相加关系,然后推导满足最大似然估计准则的窗函数,最后对乘积加权平均得到频率估计值.该方法主要优势是频率增大时依然保持较低的信噪比门限,频率越大优势越大.仿真结果表明在保证-1dB的信噪比门限前提下,估计范围仍能达到理论值的99%.进一步增强处理后,信噪比门限还能降低约1.5dB.同时满足了频率估计中的低信噪比门限和大估计范围两项要求.A new frequency estimator based on autocorrelation function was proposed. Firstly, the conjugate product of normalized sample autocorrelation function was calculated, and then the multiplic- ative relationship between signal and noise of the product was converted into additive relationship, and then a window function was derived to smooth the conjugate product to get the estimated value of fre- quency. Proposed method keeps low SNR threshold when frequency to be estimated becomes larger, while other method is frequency-sensitive at low SNRs. Simulation results show that proposed method can keep --1 dB in SNR threshold and 99% of the theory value in estimation range simultaneously. Additional enhanced process can bring extra 1.5 dB lower in SNR threshold. Propose method meets the requirements both in SNR threshold and estimation range.

关 键 词：数字通信系统 频率估计 自相关函数 信噪比门限 估计范围 最大似然估计 

分 类 号：TN911[电子电信—通信与信息系统] TN914[电子电信—信息与通信工程]