α稳定分布噪声下基于稳健S变换的LFM信号参数估计

发布时间：2018-01-11 20:24

本文关键词：α稳定分布噪声下基于稳健S变换的LFM信号参数估计　出处：《系统工程与电子技术》2017年04期 　论文类型：期刊论文

【摘要】：S变换由短时傅里叶变换发展而来,克服了短时傅里叶变换窗长固定、不能同时展现信号高频及低频的缺点,但在脉冲性较强的α稳定分布噪声下,该方法性能退化甚至失效。对此,基于广义柯西分布,构造了一类可有效应用于强脉冲噪声环境的损失函数,并详细分析了其影响函数的稳健性。在此基础上,根据最大似然估计理论和S变换,提出了一种稳健S变换方法。该方法以S变换作为初始值,采用最大似然估计方法在时频域迭代得到,在保留S变换窗长选取灵活等优点的同时,进一步提高了S变换的时频聚集性。仿真实验表明,在处理脉冲噪声环境下的线性调频信号时,与传统的基于Myriad滤波、Meridian滤波等多种非线性滤波的方法相比,提出的稳健S变换不仅能有效抑制脉冲噪声,且在脉冲性较强的α稳定分布噪声环境下,具有良好的鲁棒性和优良的线性调频信号参数估计性能。
[Abstract]:S transform and development by short time Fourier transform, short-time Fourier transform to overcome the fixed window length, cannot show the signal of high frequency and low frequency faults at the same time, but in the pulse of alpha stable distribution noise, the method of performance degradation and even failure. Therefore, based on the generalized Cauchy distribution, construct the loss function of a class can be used effectively in strong impulsive noise environment, and analyzed its influence function robustness. On this basis, according to the theory of maximum likelihood estimation and S transform, this paper presents a robust S transform method. This method is based on the S transform as the initial value, using the maximum likelihood estimation method in frequency domain iterative, while retaining the S transform window select the long flexible at the same time, to further improve the time-frequency transform S aggregation. Simulation results show that the processing of linear frequency modulation pulse signal in noise environment, and the traditional filter based on Myriad Compared with other nonlinear filtering methods such as Meridian filtering, the proposed robust S transform can not only suppress impulse noise effectively, but also has good robustness and excellent performance in LFM signal parameter estimation in a strong pulsed alpha stable distributed noise environment.

【作者单位】：西安电子科技大学电子工程学院;
【基金】：国家自然科学基金(61201286) 陕西省自然科学基金(2014JM8304)资助课题
【分类号】：TN911.23
【正文快照】： 0引言在传统的信号处理中,所分析的背景噪声干扰一般为高斯噪声。然而在工程应用中,实际噪声通常具有明显的脉冲特性,如海杂波噪声、大气噪声、无线信道噪声等。这类噪声持续时间短,幅度大,由不规则脉冲或尖峰组成,其概率密度函数拖尾严重[1-3]。由于α稳定分布是唯一一类满足

【相似文献】