基于子带分解的宽带自适应波束形成算法改进

Improvement of the broadband adaptive beamforming algorithm based on sub-band decomposition

该文针对在目标快速运动或者本船快速机动情况下,最小方差无失真响应(minimum variance distortionless response,MVDR)在少快拍下无法获得满秩和稳定的协方差矩阵的问题,提出一种基于子带分解的宽带自适应波束形成方法,该方法能够在单快拍数据条件下,获得满秩的协方差矩阵。相对于时域解析信号的MVDR算法,基于子带分解的宽带信号处理方式具有更小的相位误差,波束形成结果更精准。针对其在阵元失效情况下性能急剧下降的问题,对比分析了采用基于数据的协方差矩阵重构方法、对角加载方法、阵元隔离方法在阵元失效情况下的性能。仿真分析结果表明,基于数据的协方差矩阵重构方法,在阵元失效的情况下,具有相对较低的旁瓣高度。海试数据处理结果表明,改进的宽带自适应波束形成算法可在频域单次快拍数据条件下完成波达方向(direction of arrival,DOA)估计,且具有较高的DOA估计精度、方位分辨能力和弱目标检测能力。

Focusing on the problem that MVDR cannot obtain a full rank and stable covariance matrix with few snapshots when the target is moving rapidly or the ship is maneuvering rapidly,a broadband adaptive beamforming method based on sub-band decomposition was proposed.A full-rank covariance matrix can be obtained under the condition of single snapshot data.Compared with the MVDR algorithm that analyzes signals in the time domain,the broadband signal processing method based on sub-band decomposition has smaller phase errors and more accurate beamforming results.Aiming at the problem that the performance degrades greatly when the array element fails,the performance of the covariance matrix reconstruction method,the diagonal loading method,and the array element isolation method was compared and analyzed.The simulation analysis results show that the data-based covariance matrix reconstruction method has a relatively low sidelobe height in the case of array failure.The experimental results show that the improved broadband adaptive beamforming algorithm can complete DOA estimation under the condition of single snapshot data,and has higher DOA estimation accuracy and azimuth resolution ability.