一种新的信号子空间宽带测向算法

宽带信号测向是信号处理研究的难点课题之一，经典的旋转信号子空间（RSS）变换算法需要知道和真实来波方向接近的方向初始值。不合适的方向初始值将会降低估计器的收敛能力，引起估计偏差，从而使算法的估计精度受预估计的影响很大。本文提出的新算法首先推导出不同频段不同方向角的方向矢量之间的关系，然后再利用信号子空间、噪声子空间和阵列流型之间的关系构造一个特殊矩阵，最后通过求条件数来判断来波方向。所以新算法不需要构造聚焦矩阵，不需预知初始值，从而提高了算法的稳健性，降低了运算量，且新算法同时利用了信号子空间和噪声子空间，减少了噪声子空间扰动对算法精度的影响。计算机仿真验证了新算法的有效性。

New signal subspace direction-of-arrival estimator for wideband sources

Wideband direction finding is one of the difficult tasks in signal processing. The classical rotational signal subspace transform algorithm requires initial values which are close enough to the true values. The preprocessing procedure should be carefully done lest it introduce some distortion. Failure to find proper initial values may prevent convergence of the estimator or cause biases in the estimator, consequently influence the estimation precision. The proposed method educes the relation of the different frequency and direction bearing vector, then constructs an especial matrix by use of the connection of signal subspace, noise subspace and array manifold, finally estimates DOA via condition number. Therefore it does not require any initial values, thus can improve the steadiness of new algorithm and reduce the calculation cost. By use of the signal subspace and the noise subspace of decomposed wideband signals, the influence of noise subspace disturbance on estimation precision is reduced. Simulation results verified the efficiency of the new method.