冲击噪声下基于相关熵的二维ESPRIT算法  

Correntropy-Based Two-Dimensional ESPRIT Algorithm under Impulsive Noise

作　　者：赵季红[1,2] 李雷雷[1] 曲桦[2] 王伟华[2] 马兆恬 

机构地区：[1]西安邮电大学通信与信息工程学院,西安710061 [2]西安交通大学电子信息学院,西安710049

出　　处：《计算机测量与控制》2016年第6期170-173,184,共5页Computer Measurement &Control

基　　金：国家自然科学基金项目(61371087);新一代宽带无线移动通信国家重大专项项目(2013ZX0302010-003);国家高技术研究发展计划("863"计划");(2014AA01A706)

摘　　要：针对基于相关函数的波达方向(direction of arrival,DOA)估计方法在冲击噪声环境下性能下降明显甚至失效的问题,提出基于相关熵(Correntropy)的二维ESPRIT算法;该方法利用相关熵在冲击噪声环境下具有鲁棒性的优点,将受干扰信号的自相关函数替换为相关熵函数,并结合二维ESPRIT算法实现在冲击噪声环境下进行二维DOA估计;仿真表明,与基于分数低阶统计(Fractional Lower Order Statistics,FLOS)算法相比,该算法呈现明显优势,特别在高的冲击噪声条件下(1<α<1.5)能对信源方向进行更加有效的估计,且均方误差值仍保持很低。In order to solve the problem that the correlation function-based DOA （Direction of Arrival,DOA）estimation method has an obvious performance degradation or even a total efficiency loss in the impulsive noise environment,a new two-dimensional ESPRIT algorithm based on correntropy was proposed.It takes advantage of correntropy’s robustness in the impulsive noise environment,replaces the signal＆#39;s autocorrelation function with the correntropy function and combines with two-dimensional ESPRIT algorithm to estimate two-dimensional DOA in the impulsive noise environment.Simulation results show that CRCO-ESPRIT outperforms the existing FLOS algorithm and espe-cially can estimate DOA efficiently in high impulsive noise environments （1＆lt;α＆lt;1. 5）,and the mean square error still remain at a low level.

关 键 词：相关熵 波达方向估计 冲击噪声 分数低阶统计 二维ESPRIT算法 

分 类 号：TN911[电子电信—通信与信息系统]