Fast Forward-Backward Hankel Matrix Completion for Automotive Radar DOA Estimation Using Sparse Linear Arrays
Fast Forward-Backward Hankel Matrix Completion for Automotive Radar DOA Estimation Using Sparse Linear Arrays
复制标题
DOI：
10.1109/radarconf2351548.2023.10149466
复制
发表时间：
2023-05
期刊：
影响因子：
--
通讯作者：
Shunqiao Sun;Yining Wen;Ryan Wu;D. Ren;Jun Li
中科院分区：
文献类型：
--
作者：
Shunqiao Sun;Yining Wen;Ryan Wu;D. Ren;Jun Li
具有稀疏线性阵列的汽车多输入多输出 (MIMO) 雷达是一种经济高效的解决方案，可以以较低的硬件成本和减少的互耦实现大孔径尺寸。与汽车 MIMO 稀疏线性阵列相关的挑战是高旁瓣，这可能会导致角度检测误差。本文提出了一种利用 Hankel 矩阵结构进行联合阵列插值和超分辨率单快照测角的快速前向-后向 Hankel 矩阵补全和矩阵铅笔方法。所提出方法的新颖性在于两个部分。它不仅节省了矩阵补全每次迭代中奇异值分解（SVD）的计算成本，而且增加了使用相同数量天线单元构建更大维数低秩矩阵的自由度，从而可以以更高的精度完成和估计更多的目标。数值结果证明了该方法的有效性和效率。
Automotive multiple-input multiple-output (MIMO) radar with sparse linear arrays is a cost-effective solution to achieve large aperture size with low hardware cost and reduced mutual coupling. The challenges associated with automotive MIMO sparse linear arrays are the high sidelobes, which might result in angular detection errors. This paper presents a fast forward-backward Hankel matrix completion and matrix pencil method for joint array interpolation and super-resolution single-snapshot angle finding by exploiting the structure of Hankel matrix. The novelty of the proposed approach lies in two parts. It not only saves the computational cost of singular value decomposition (SVD) in each iteration of the matrix completion, but also increases the degrees of freedom to construct a low-rank matrix with larger dimensions using the same number antenna elements, as a result of which, more targets can be completed and estimated with better accuracy. Numerical results demonstrate the effectiveness and efficiency of the proposed method.