非均匀噪声背景下的欠定DOA估计方法

摘要

针对非均匀噪声背景下欠定波达方向(direction of arrival,DOA)估计问题,结合互质阵列的结构优势,提出了基于全变分范数最小化的DOA估计方法.首先利用连续差联合阵列与连续波程差一一对应的特性,构造出新的阵列接收数据,阵列孔径得到扩展;然后将其转化为一个联合优化问题,在代价函数中利用全变分范数和 L1范数惩罚项分别对角度的稀疏性和噪声项进行约束;最后通过求解相应的凸优化问题以及多项式求根得到DOA的高精度估计.与现有方法相比,所提方法不仅无需进行预白化处理,而且考虑了连续角度域内的所有角度信息而不是对角度域进行离散化,有效避免了模型失配对估计性能的影响,提高了估计精度和分辨率.仿真实验验证了所提方法的有效性与优越性.%For the problem of underdetermined direction of arrival(DOA)estimation with nonuniform noise,a DOA estimation method is proposed with coprime array based on total variation norm minimization. Firstly,received data is constructed by using the property that the consecutive difference coarray is corre-sponding to the consecutive wavepath difference,and the array aperture is extended.Then,a joint optimiza-tion based on total variation norm is constructed,and add a total variation norm penalty on the sparsity of an-gle and L1norm penalty on noise.Finally,the estimation of DOA is achieved by solving the corresponding convex optimization problem and polynomial rooting.Compared with the existing methods,the proposed method not only does not need the pre-w hitening process but also considers a continuous range of possible DOA instead of discretizing this range onto a grid,off-grid effects can be neglected,thus precision and reso-lution of DOA estimation can be improved.Extensive simulations show that the effectiveness and superiority of the proposed method.