利用空域稀疏性的L型阵下二维波达方向估计

摘要

The main difficulty of casting sparse reconstruction into two dimension direction-of-arrival (2D-DOA) estimation is the heavy computational burden.In order to overcome this problem,we propose a novel 2D-DOA estimation method in terms of reducing the number of atoms and the model order in the sparse linear model to be solved.In proposed method,the traditional sparse linear model based on array received data is decomposed,transformed and recombined according to the characteristic of configuration in L-structure array.This procedure discards those inactive atoms unrelated to sources and results in an obvious decrease in the number of atoms.Furthermore,based on the eigenvector corresponding to the maximal eigenvalue of covariance matrix of array received data,a new sparse linear model,whose model order is one,is built and substitutes for the traditional sparse linear model.Moreover,the detailed steps of the proposed approach are present and the computational complexity is also analyzed.At last,the excellent estimation performance of our method is demonstrated through a series of computer simulations.%针对稀疏重构应用于二维波达方向估计时存在计算量大的问题,从减少过完备字典的原子个数以及求解模型的阶数这两个方面入手,提出一种具有较低计算复杂度的基于稀疏重构的二维波达方向估计方法.该方法中首先根据L型阵的结构特点,对基于阵列接收数据的稀疏线性模型进行分解和重新组合,抛弃与信源无关的原子,大大减少了原子个数；其次,提出利用多快拍数据协方差矩阵的最大特征值对应的特征向量作为待分解信号建立稀疏模型,使模型阶数从快拍数降至一阶.此外,给出了新方法的详细实施步骤,并且对其计算复杂度进行了理论分析.新方法估计精度高,并且能够自动实现俯仰角和方位角配对.最后,通过计算机仿真验证了新方法的有效性.