基于循环平稳分析的LFM信号盲处理结果可靠性评估  

Reliability Evaluation for Blind Processing Results of LFM Signal Based on Cyclostationarity

作　　者：胡国兵[1,2] 徐立中[2] 吴珊珊[1] 高燕[1] 王书旺[1] 

机构地区：[1]南京信息职业技术学院电子信息学院,江苏南京210023 [2]河海大学计算机与信息学院,江苏南京210098

出　　处：《电子学报》2016年第4期788-794,共7页Acta Electronica Sinica

基　　金：江苏省基础研究计划(自然科学基金)(No.BK2011837);江苏省"333高层次人才培养工程"(No.BRA2013171);江苏省政府留学基金(No.JS-2007-105);江苏省"青蓝工程";江苏省"六大人才高峰"第十二批高层次人才资助项目(No.DZXX-022)

摘　　要：针对线性调频(LFM,Linear Frequency Modulation)信号盲处理结果的可靠性评估问题,提出了一种基于循环频率特征分析的处理算法.首先对观测信号进行调制方式识别及参数估计,并据此建立参考信号,后将观测信号与参考信号作相关运算.通过检测相关序列在零频率附近是否存在循环频率,实现对LFM信号盲处理结果的可靠性检验.文中对所提出检验算法的错误概率进行了理论推导,并以常用的离散多项式变换(DPT,Discrete Polynomial Transform)法为例进行了实证分析.仿真结果表明,相对于已有时域方法而言,本文算法无需估计信噪比,且在低信噪比条件下具有更好的统计性能.A reliability test method based on cyclic frequency features was proposed to evaluate the blind processing results of linear frequency modulation （LFM） signal. The reference signal was constructed depending on the identified mod- ulationtype and the correspondent parameters estimations at first and the correlations between the observed signal and the ref- erence signal were calculated consequently. The reliability for LFM signal processing result was tested by detecting the pres- ence of cyclic frequencies near zero frequency of the correlations. The theoretical error probability of the detection was de- rived and the case study for the discrete polynomial transform （DPT） method was described in the paper. Simulation results show that the proposed algorithm avoids estimating of signal-to-noise ratio and is superior to the time domain based method under lower signal-to-noise ratios.

关 键 词：盲信号处理 可靠性检验 循环频率 线性调频信号 

分 类 号：TN911.6[电子电信—通信与信息系统]