基于噪声子空间估计的MWF实现  

Implementation of MWF Based on Noise Subspace Estimation

作　　者：周柱[1] 张尔扬[1] 卢树军[1] 

机构地区：[1]国防科技大学电子科学与工程学院,长沙410073

出　　处：《宇航学报》2012年第5期661-668,共8页Journal of Astronautics

基　　金：通信抗干扰技术国家级重点实验室基金(9140CD202011003)

摘　　要：采用阵列天线对GPS接收信号进行干扰抑制,在信号处理时引入时延,形成空时二维处理的模式。空时二维抗干扰由于运算量大而导致在实际实现困难,必须进行降维处理。多级维纳滤波(MWF)方法可以有效降低滤波器的维数,但是经典的MWF方法存在子空间维数估计不准的问题。本文对于多级维纳滤波方法进行了分析,利用MWF的分析滤波器将接收信号矢量映射为另一信号矢量。通过对该信号矢量的协方差矩阵进行分析,找到一种判断子空间维数的稳健方法。仿真表明该方法能够准确地估计出噪声子空间维数。与传统的设定MSE门限的方法相比较,得出用本方法估计子空间维数更为准确可靠,抗干扰性能更优的结论。The interference suppression of GPS received signal is realized by utilizing array antenna in this paper, and time delay is introduced for signal processing, thus forming a spatial-time two dimension processing mode. To realize the two dimensional processing on interference suppression, the large amount of operation is a difficulty, so it is necessary to reduce the rank. Multistage Weiner Filter （MWF） is an effective way to reduce the rank of the filter, but a drawback exists, that the subspace rank of noise can＇ t be estimated with the classical MWF method. The MWF is analyzed in this paper, and the analytical part of MWF is adopted to map the received signal into a new signal vector. Through analyzing the eovariance matrix of the new signal vector, a robust method is proposed to estimate the subspace rank. Simulations show that the subspace rank can be estimated accurately by the proposed method. Comparing the proposed method with the traditional method of setting the threshold of MSE, the subspaee rank estimated is more accurate and credible, and it exhibits superior anti-jam performance.

关 键 词：GPS 多级维纳滤波 秩选 MDL 二分法 

分 类 号：TN911.72[电子电信—通信与信息系统]