针对宽带多相干组信号条件下传统的聚焦类算法性能下降的问题,提出一种基于空间斜投影的宽带波达方向(DOA)估计算法。该算法首先利用窄带解相干算法得到初始的DOA估计值,接着通过求解约束最小化问题得到各相干组信号的衰落系数估计值;然后,利用空间斜投影技术将不同的相干组信号分离,并利用聚焦变换将分离的信号聚焦到单一频点处;最后,利用窄带高分辨算法得到最终的DOA估计值。仿真实验表明:与现有的聚焦类算法相比,文中算法具有更好的估计性能,在较低信噪比情况下尤为明显;同时,文中方法对于相关性较低的空间色噪声也有一定适用性。

According to the problem that the traditional focusing methods have a performance decline under conditions of multiple groups of wideband coherent signals, a novel method based on subspace oblique projection is proposed. The traditional narrowband decorrelation algorithm is firstly used to estimate the initial DOAs. Then the fading coefficients of different groups of wideband coherent signals can be estimated by solving a constrained minimization problem. Afterwards, signals that belong to different groups are separated by means of subspace oblique projection. The separated signals are focused onto a single frequency by means of focusing transformation. At last, the narrow-band high resolution algorithms are utilized to get the final DOA estimation. Simulation results show that the proposed algorithm has a better estimation performance compared with the existing focusing algorithms especially with a lower signal-to-noise ratio. And this algorithm also has certain applicability for spatially colored noise with lower correlation.

TN911.7

国家自然科学基金资助项目