论文部分内容阅读
对信号非圆特性的有效利用能显著改善子空间类阵列测向方法的性能,但难以弥补此类方法在低信噪比(SNR)、小样本等信号环境适应能力方面的局限。本文引入贝叶斯稀疏学习(SBL)技术以解决非圆信号的波达方向(DOA)估计问题,在结合信号非圆特性的同时对入射信号的空域稀疏性加以利用,通过将非圆信号阵列输出协方差矩阵和共轭协方差矩阵在预先定义的空域字典集上进行稀疏重构,得到入射信号的空间谱重构结果,并依据其谱峰位置估计各信号的方向。该方法对独立和相关信号都具有较好的适应能力,仿真结果验证了该方法在信号环境适应能力和相关信号测向精度等方面的优势。
The effective use of signal non-circular features can significantly improve the performance of subspace-based array DF method, but it is difficult to make up for the limitations of such methods in low signal-to-noise ratio (SNR) and small sample signal adaptability. In this paper, Bayesian Sparse Learning (SBL) is introduced to solve DOA (DOA) estimation of non-circular signals. The spatial sparsity of incident signals is used in combination with non-circular features of signals. The output covariance matrix and the conjugate covariance matrix are sparsely reconstructed on a predefined airspace dictionary set to obtain the spatial spectrum reconstruction result of the incident signal, and the direction of each signal is estimated according to the peak position of the signal. The proposed method has good adaptability to both independent and correlated signals. The simulation results verify the advantages of this method in signal environment adaptability and signal-to-signal accuracy.