基于压缩感知的米波雷达低空测角算法

Low-angle estimation method based on compressed-sensing for meter-wave radar

为解决多径环境下米波雷达对低空目标探测问题,本文结合信号空域稀疏性和多径模型下的复合导向矢量提出了一种多径环境下高分辨的低空测角方法,能够有效地克服多径效应问题。该方法首先利用多径模型下的先验信息产生复合导向矢量,然后利用该导向矢量构造压缩感知矩阵,此时的感知矩阵是整合了多径衰减系数和回波角度关系等先验信息,同时通过对多快拍数据矩阵的奇异值分解获得较高信噪比的信号数据矩阵,继而利用感知矩阵和信号数据矩阵建立最优化L1范数约束求解模型,最后利用凸优化工具求解稀疏空间谱,估计直达波和反射波入射角度值。该方法能够增强信息矢量稀疏性,在较低信噪比下可获得高分辨的角度估计性能。仿真实验证明了该方法的优越性。

For the low angle estimation problem of meter-wave radars, based on the signal spatial sparseness and the complex steering vector in muhipath, a new high-resolution angle estimation method is proposed. First, this method utilizes priori information of the muhipath model to produce complex steering vector, and then the compressed sensing matrix is constructed, which contains priori information such as muhipath attenua- tion coefficient and angle relationship. Meanwhile, by singular value decomposition （SVD） of the multi snap data matrix, a data matrix with higher signal-noise ratio （SNR） can be obtained. Afterwards, the compressed sensing matrix and the new data matrix are used to construct the optimal L1-norm constraint model. Finally, convex optimality tools are used to calculate the sparse spectral function. This method can improve the sparse- ness of information vector and it has a high-resolution performance at low SNR in multipath. Simulation results demonstrate the advantage of the proposed method over the L]-SVD method.