基于修正的子空间正交测试宽带DOA估计方法

A New Wideband Direction-of-arrival Estimation Method using Modified TOPS Algorithm

投影子空间正交性测试法(TOPS)是利用宽带信号多个频点的噪声子空间与信号子空间的正交性实现到达角(DOA)估计。在中等信噪比时该方法估计性能较好,而在其他信噪比条件下空间谱易存在多个"伪峰",算法性能依赖于参考频点的选择。针对该问题,提出了一种新的TOPS算法,该方法通过最大化各频率点信号子空间与噪声子空间特征值区分度选择参考频点,同时利用信号子空间投影代替其零空间投影,避免信号子空间估计误差导致空间谱中产生伪峰,最后利用子空间的正交性实现宽带DOA估计。仿真结果表明,相比于传统的TOPS算法,该方法在一定信噪比条件下,避免了伪峰的出现,提高了TOPS算法的估计精度和分辨率。

The test of orthogonality of projected subspaces （TOPS）estimates directions of arrival of wideband sources by measuring the orthogonal relation between the signal and the noise subspaces ofmultiple frequency components of the sources. TOPS performs well at mid signal-to-noise ratio（SNR）range,but fares poorly at other SNR cases. The TOPS pseudo spectrum often exhibits spurious peaks atall SNR levels Depending on which frequency bins are chosen,the estimator performance could bedifferent. In this paper,the suitable modifications to extend the effectiveness of TOPS is proposed.First,we may choose the signal subspace of the frequency bin where the separation between thesmallest eigenvalue in the signal subspace and the largest eigenvalue in the noise subspace ismaximum. The proposed modified -TOPS （MTOPS） achieves reduction in spurious peaks by incorporating signal -subspace projection instead of null -space projection used in TOPS. The performance of MTOPS is compared with that of original TOPS through computer simulations. The simulations show that the proposed MTOPS achieves reduction in spurious peaks and performs betterthan original TOPS in some signal-to-noise ratio（SNR）ranges.