利用RR-MWF的低信噪比下机载气象雷达回波谱矩估计方法  

A RR-MWF Based Spectral Moments Estimation Algorithm for Airborne Weather Radar Echoes with Low SNR

作　　者：卢晓光[1] 范源丹 李海[1] 吴仁彪[1] Lu Xiaoguang;Fan Yuandan;Li Hai;Wu Renbiao(Tianjin Key Lab for Advanced Signal Processing,Civil Aviation University of China,Tianjin 300300,China)

机构地区：[1]中国民航大学天津市智能信号与图像处理重点实验室

出　　处：《信号处理》2019年第7期1125-1132,共8页Journal of Signal Processing

基　　金：国家自然科学基金(U1633106,61471365,U1733116)

摘　　要：常规机载气象雷达采用脉冲对法估计谱宽实现湍流检测,当信噪比较低时脉冲对法的谱宽估计误差大。晴空湍流(clear air turbulence,CAT)含水量较少,雷达回波信噪比很低,因此常规机载气象雷达无法检测CAT。为提高低信噪比下机载气象雷达回波谱宽估计性能,提出了一种基于降秩多级维纳滤波器(Reduced-Rank Multistage Wiener Filter,RR-MWF)的回波谱矩估计方法。该方法在机载气象雷达引入空时体制的基础上,利用空时域联合处理对湍流回波进行处理,通过空时积累改善信噪比。在最小均方误差准则下,构造了适用于分布式气象目标的自适应RR-MWF权矢量和代价函数,估计回波谱矩。仿真实验表明,提出的RR-MWF估计器在信噪比低于10 dB时明显优于常规的脉冲对法,可用于CAT检测。The precipitation of CAT(Clear Air Turbulence)is lower than that of the convective induced turbulence,resulting echoes with low SNR are received by classic airborne weather radars.The error of the spectrum width estimation is large in low SNR scenarios where the widely used pulse pair processing(PPP)method for airborne weather radars is used,therefore the PPP method can’t be used for detecting CAT.A spectral moments estimation method based on reduced-rank multistage wiener filter(RR-MWF)is proposed for improving the estimation performance in low SNR scenarios.The proposed method is in essential a space-time adaptive processing algorithm,which can improve the SNR of radar echoes by integrating coherently both in the spatial and the temporal dimension.Considering that the CAT is a kind of distributed target,the adaptive RR-MWF weighted vector is newly constructed and the cost function used for estimating the spectral moments is deduced under the mean square error criterion,which are the main elements of the work.Experimental simulations and numerical analysis show that the RR-MWF estimator outmatches the PPP method in low SNRs,typically lower than 10 dB.And RR-MWF estimator can be used to detect CAT.

关 键 词：机载气象雷达 晴空湍流 低信噪比 谱矩估计 降秩多级维纳滤波 

分 类 号：TN959.4[电子电信—信号与信息处理]