基于梯度自适应格型滤波的外辐射源雷达多径杂波抑制算法  

Multipath clutter mitigation for passive radar based on gradient adaptive lattice filter

作　　者：方亮[1] 万显荣[1] 易建新[1] 饶云华[1] 柯亨玉[1] 

机构地区：[1]武汉大学电子信息学院,湖北武汉430072

出　　处：《系统工程与电子技术》2013年第11期2291-2296,共6页Systems Engineering and Electronics

基　　金：国家自然科学基金(61331012;61371197;41074116;61271400);高等学校博士学科点专项科研基金(博导类;20120141110077);武汉大学研究生自主科研项目基金(274763)资助课题

摘　　要：外辐射源雷达面临严重的多径杂波问题,传统的梯度自适应格型(gradient adaptive lattice,GAL)滤波是一种有效的杂波抑制算法,但该算法需要根据经验合理设置步长参数,收敛速率慢,在抑制杂波的同时会造成处理距离元内的目标回波被部分相消。基于传统的GAL滤波提出了一种改进的GAL算法及其前后向预测误差的归一化形式,该算法通过最小化前向和后向预测误差的均方值估计反射系数,并通过最小化多重回归滤波的均方误差估计其权系数。相对传统的算法而言,改进的算法不需设置步长参数、收敛速率快、对处理距离元内的目标回波基本没有影响。仿真和实测数据处理结果验证了该算法的有效性。While the passive radar faces a significant problem in multipath clutter, the traditional gradient adaptive lattice （GAL） filter algorithm can be efficient for multipath clutter mitigation, but an appropriate step parameter must be set with the empirical rule in the algorithm with a slow convergence rate. A partial cancella- tion of the target echoes whose delays are included in the cancelled area is even yielded while the multipath clut- ter is rejected. A modified GAL filter algorithm is proposed based on the traditional GAL filter algorithm. An alternative form of the algorithm is obtained by introducing the normalized forward and backward prediction er- rors. The basic principle of the algorithm is to select an estimation for the reflection coefficient by minimizing the sum of the mean-square forward and backward prediction errors and an estimation for the coefficient of the multiple regression filter by minimizing the mean-square error. The modified algorithm does not eed a step pa- rameter, yet exhibits relatively faster convergence rate and has little effect on the target echoes within the can- celled area compared with the traditional algorithm. The simulation and real-life data results confirm the effec- tiveness of the proposed algorithm.

关 键 词：外辐射源雷达 多径杂波抑制 梯度自适应格型滤波 

分 类 号：TN958.97[电子电信—信号与信息处理]