存在幅相误差下的稳健稀疏贝叶斯二维波达方向估计  

Robust Sparse Bayesian Two-dimensional DOA Estimation with Gain-phase Errors

作　　者：王绪虎 金序 侯玉君 张群飞[3] 徐振华[2] 王辛杰 陈建军 WANG Xuhu;JIN Xu;HOU Yujun;ZHANG Qunfei;XU Zhenhua;WANG Xinjie;CHEN Jianjun(School of Information and Control Engineering,Qingdao University of Technology,Qingdao 266520,Shandong,China;Key Laboratory of Ocean Circulation and Waves,Institute of Oceanology,Chinese Academy of Sciences,Qingdao 266071,Shandong,China;School of Marine Science and Technology,Northeast Polytechnical University,Xi'an 710072,Shaanxi,China)

机构地区：[1]青岛理工大学信息与控制工程学院,山东青岛266520 [2]中国科学院海洋研究所海洋环流与波动重点实验室,山东青岛266071 [3]西北工业大学航海学院,陕西西安710072

出　　处：《兵工学报》2024年第10期3608-3618,共11页Acta Armamentarii

基　　金：国家自然科学基金项目(62171247);山东省自然科学基金项目(ZR2021QF113、ZR2022MF273)。

摘　　要：为减小传感器幅相误差的影响,提升方位估计性能,针对L型传感器阵列提出一种存在幅相误差下的稳健稀疏贝叶斯二维波达方向(Direction-Of-Arrival,DOA)估计方法。引入一个辅助角,将二维DOA估计问题转化为两个一维角度估计问题。利用L型阵列两子阵数据互协方差矩阵的对角线元素向量,构造一个含有幅相误差的稀疏表示模型,采用期望最大算法推导未知参数表达式并进行迭代运算,进而获得离网格和信号精度,利用二者构建新的空间谱函数,通过谱峰搜索估计出辅助角;将求得辅助角代入含有幅相误差的阵列接收数据稀疏表示模型,再次运用稀疏贝叶斯学习方法,估计出入射信号的俯仰角;根据3个角之间的关系,估计出方位角。研究结果表明:该方法实现了方位角和俯仰角的自动匹配,进一步克服了幅相误差对估计性能的影响,提高了方位估计的精度和角度分辨力,尤其是在高信噪比和幅相误差较大情况下优势更明显;仿真结果验证了该方法的有效性。To reduce the influence of gain-phase errors and improve the performance of direction-of-arrival(DOA)estimation,a robust sparse Bayesian two-dimensional DOA estimation method with gain-phase errors is proposed for the L-shaped sensor array.In the proposed method,an auxiliary angle is introduced to transform a 2D DOA estimation problem into two 1D angle estimation problems.A sparse representation model with gain-phase errors is constructed by using the diagonal element vector of the cross-covariance matrix of two submatrices of L-shaped sensor array.The expectation maximization algorithm is used to derive the unknown parameter expression,which is used to perform the iterative operations for obtaining the off-grid and the precision of signal.A new spatial spectral function is constructed by using the off-grid and the precision of signal.The auxiliary angle can be estimated by searching the new spatial spectra peak.The estimated auxiliary angle is introduced into the sparse representation model of the received data with gain-phase errors,and then the sparse Bayesian learning method is used to estimate the elevation angle of incident signal.According to the relationship among three angles,the azimuth angle can be estimated.The results show that this method realizes the automatic matching of azimuth angle and elevation angle,and improves the accuracy of DOA estimation and angle resolution.Simulated results verify the effectiveness of the proposed method.

关 键 词：波达方向估计 幅相误差 稀疏信号重构 稀疏贝叶斯学习 L型阵列 

分 类 号：TN911.7[电子电信—通信与信息系统]