稀疏表示框架下无需参数配对的二维到达角估计

现有二维到达角估计算法大多基于子空间理论及需要参数配对,针对这一问题,在稀疏表示理论框架下提出了一种参数自动配对的二维到达角估计新算法。该算法在L阵列下构建阵列互相关矩阵的稀疏表示模型,利用奇异值分解降低复杂度并基于群LASSO(Least Absolute Shrinkage and Selection Operator)获得方位角估计。在方位角估计的基础上,基于向量化操作构建稀疏空间谱匹配模型,然后利用LASSO 获得俯仰角估计。与参数配对ESPRIT 和改进的传播算子方法相比,所提算法不仅无需参数配对过程,而且可以提供改进的估计精度。计算机仿真结果验证了所提算法的有效性。

2-D Angle of Arrival Estimation without Pair Matching under Sparse Representation Framework

Most of existing 2-dimensional (2-D) angle of arrival (AOA) estimation algorithms rely on subspace technique and require pair matching process. This paper presents a novel and automatically paired 2-D AOA estimation algorithm in sparse representation framework. The algorithm constructs a sparse representation model of the cross-correlation matrix under L-shaped array and reduces the computational complexity by singular value decomposition (SVD), then obtains azimuth angle estimation via group LASSO. Based on the estimate result of azimuth angle, a sparse spatial spectrum fitting model is generated by vectorization operator, and elevation angle estimation is achieved successively via LASSO. Compared with pair-matching ESPRIT and modified propagator method, the proposed algorithm not only avoids parameter pair matching process, but also provides improved estimation accuracy. The simulation results validate the effectiveness of the proposed algorithm.