低秩条件下的波达方向估计方法研究
发布时间：2018-08-03 07:26
【摘要】：在波达方向估计中,阵列接收数据(观测矩阵)的秩小于信源数时会导致经典的基于子空间分解类的超分辨算法失效,基于压缩感知理论的超分辨方法虽可解决算法失效问题,但随着观测矩阵秩的增加,压缩感知方法大多只是进行简单的向量范数合成,即将多测量向量问题转变到单次测量向量问题来解决,并没有充分利用多余的采样信息.在研究秩缺失条件下信号重构机理的基础上,提出了一种自适应加权递归算法,能够利用额外的采样信息通过空间投影构造出相对正确的信号子空间,弥补了在秩缺失情况下估计精度差的问题,并且在采样数逐渐增加的基础上,可以实现对信号的无偏估计.
[Abstract]:In DOA estimation, when the rank of array received data (observation matrix) is less than the number of sources, the classical super-resolution algorithm based on subspace decomposition class will fail, and the super-resolution method based on compressed sensing theory can solve the problem of algorithm failure. However, with the increase of the rank of observation matrix, compression sensing methods are mostly used in simple vector norm synthesis, that is, the multi-measurement vector problem is transformed to the single-measurement vector problem to solve the problem, and the redundant sampling information is not fully utilized. On the basis of studying the mechanism of signal reconstruction under the condition of lack of rank, an adaptive weighted recursive algorithm is proposed, which can make use of extra sampling information to construct a relatively correct signal subspace by spatial projection. It makes up for the problem of poor estimation accuracy under the condition of lack of rank, and the unbiased estimation of signal can be realized on the basis of increasing the number of samples.
【作者单位】： 海军工程大学电子工程学院;哈尔滨工业大学电子与信息工程学院;
【分类号】：TN911.23
[Abstract]:In DOA estimation, when the rank of array received data (observation matrix) is less than the number of sources, the classical super-resolution algorithm based on subspace decomposition class will fail, and the super-resolution method based on compressed sensing theory can solve the problem of algorithm failure. However, with the increase of the rank of observation matrix, compression sensing methods are mostly used in simple vector norm synthesis, that is, the multi-measurement vector problem is transformed to the single-measurement vector problem to solve the problem, and the redundant sampling information is not fully utilized. On the basis of studying the mechanism of signal reconstruction under the condition of lack of rank, an adaptive weighted recursive algorithm is proposed, which can make use of extra sampling information to construct a relatively correct signal subspace by spatial projection. It makes up for the problem of poor estimation accuracy under the condition of lack of rank, and the unbiased estimation of signal can be realized on the basis of increasing the number of samples.
【作者单位】： 海军工程大学电子工程学院;哈尔滨工业大学电子与信息工程学院;
【分类号】：TN911.23
【参考文献】
相关期刊论文 前2条
1 丁君,王s，
本文编号：2161034
本文链接：https://www.wllwen.com/kejilunwen/wltx/2161034.html
