基于最大特征向量的改进TOPS算法

Improved TOPS Algorithm Based on Maximum Eigenvector

宽带投影子空间正交性测试法(TOPS)算法利用带宽内多个频点的信号子空间和噪声子空间的正交性完成波达方向(Direction of Arrival,DOA)估计。但该方法对参考频点较为敏感,不合适的参考频点会带来伪峰,且TOPS算法需要对宽带信源数目进行预估。为解决此难题,提出一种改进算法。该方法通过对各频点的阵列协方差矩阵进行特征分解,选择最大特征值对应的向量矩阵作为估计的Signal Subspace,无需信源数目的先验信息就可实现宽带信号的DOA估计。改进算法同时也克服了TOPS算法只适合中等信噪比的缺点,算法稳健性高,通用性好,仿真结果验证了算法的有效性。

The test of orthogonality of projected subspaces （TOPS） utilizes the orthogonality of signalsubspace and noise subspace of multiple frequency points of wideband signal to complete the directionsof arrival （DOA） estimation. This method is sensitive to the reference frequency point, and the unsuitablereference frequency will bring the pseudo-peak, and the TOPS algorithm needs to estimate the number ofwideband sources. In order to solve the problem, an improved algorithm is proposed. By decomposing thearray covariance matrix of each frequency point, the vector matrix corresponding to the maximum eigen鄄value is selected as the estimated signal subspace, and the DOA of the broadband signal can be realizedwithout the prior information of the number of sources Estimation. The new algorithm also overcomes theshortcoming of TOPS algorithm and has high robustness and good generality. The computer simulationconfirms the effectiveness of the method.