基于SRM模型的宽带MVDR快速算法研究  

Research on MVDR fast algorithm for wide broadband signal based on spiked random matrix model

作　　者：何宪文 温东阳 董永峰[1] 李智忠[1] HE Xian-wen;WEN Dong-yang;DONG Yong-feng;LI Zhi-zhong(Naval Submarine Academy,Qingdao 266199,China;Naval Radar Sonar Repair Factory,Qingdao 266100,China)

机构地区：[1]海军潜艇学院,山东青岛266199 [2]海军雷达声呐修理厂,山东青岛266100

出　　处：《舰船科学技术》2024年第8期126-129,共4页Ship Science and Technology

摘　　要：为满足潜艇等水下航行器宽带自适应波束形成算法的更高实时性要求,在经典的MVDR(Minimum Variance Distortionless Response)时域解析信号模型的基础上,设计一种基于SRM(Spiked Random Matrix)的协方差逆运算的快速估计算法。首先对宽带信号进行时域解析信号的构造,依据基于随机矩阵理论,然后通过对快拍信号的协方差矩阵的特征值进行调整和处理,实现精确的协方差矩阵的逆运算估计,避免了传统的复杂求逆运算,尤其适合多传感器和快拍样本受限的环境。仿真实验和海试数据证明,提出的算法波束指向性能更优,信号处理复杂度更低,随着阵元间距数的增加,性能获得明显的提升。In order to satisfy the instantaneity requirement of adaptive beamforming on broadband signal for submarine or other underwater vehicle,the fast estimation algorithm for inverse operation of covariance matrix base on spiked random matrix,which is on the basis of classical MVDR beamforming algorithm with time-domain analysis signal model.First,we construct the time-domain analysis on received broadband signal,moreover,based on random matrix theory,the eigenvalues of covariance matrix of snapshot signalare adjusted and processed to estimate inverse operation of covariance matrix precisely,which avoid the complex conventional inverse operation especially suit on the conditions for multiple sensor and snapshot limitation.Simulations and sea trials prove that the proposed algorithm has the better directivity and less complexity.With increasing number of spacing between elements,the performance is improved obviously.

关 键 词：宽带信号 SRM模型 MVDR 快速算法 

分 类 号：TN911[电子电信—通信与信息系统]