相关色噪声下无冗余累积量稀疏表示DOA估计

针对传统均匀线阵中四阶累积量计算复杂度大、对快拍数敏感的问题,提出了一种快速去冗余的高分辨波达方向估计新方法。该方法首先通过构造选择矩阵对四阶累积量矩阵进行第1次降维处理,摒弃传统四阶累积量中大量冗余数据,然后对无冗余累积量矩阵进行矢量化并通过二次降维得到统计性能更优的向量观测模型,最后在相应的过完备基下建立观测模型的稀疏表示进行波达方向(Direction of Arrival,DOA)估计。同时将方法推广到L型阵列2维DOA估计,扩展了其应用范围。与传统的四阶累积量方法相比,该方法大大地减小了计算量,对快拍数要求不高,并且能够有效地抑制相关色噪声。理论分析和仿真实验验证了该方法对1维和2维DOA估计都具有较高的估计精度和分辨率。

DOA estimation using non-redundant cumulants sparse representation in correlated colored noise

This paper aims at the problem that conventional fourth-order cumulants have high computational complexity and are sensitive to data samples.A new direction of arrival (DOA) estimation method is proposed to eliminate the redundancy quickly.The massive redundant data are removed by selection matrix to reduce the dimension of fourth-order cumulants matrix.By vectorizing the non-redundant cumulants matrix and reducing the dimension again,the vector measurement model with better statistical performance is then obtained.The sparse representation of the measurement model corresponding to the related over-complete basis is constructed for DOA estimation.Then the method is extended to L array for two dimensional DOA estimation.Compared with conventional fourth-order cumulants methods,the proposed method can greatly reduce the computational complexity and the impact of the size of data samples,and can efficiently suppress correlated colored noise.Theoretical analysis and simulation experiments verify that the proposed method has higher resolution and better estimation accuracy for one and two dimensional DOA estimation.