基于虚拟阵列的非圆信号DOA估计

Virtual Array Based DOA Estimation of Non-circular Signals

针对传统的子空间类波达方向(Direction of Arrival,DOA)估计算法只适用于入射信号个数少于天线数的局限性,利用现代通信系统中常用的非圆信号实值特性,提出了一种虚拟阵列多重信号分类法(Virtual Array Based Multiple Signal Classification,VA-MUSIC)。该方法通过对阵列输出信号进行共轭重构和合并,获得虚拟阵列来增加阵列的有效孔径。更进一步,结合空间平滑技术有效地解决了相干信号的DOA估计问题。与传统的MUSIC算法相比,新算法不仅可以增加最大可估计信源数,而且在DOA估计精度、信号源角分辨能力等方面均有明显的优势。计算机仿真验证了该算法的有效性和优越性。

The classical direction of arrival( DOA) estimation method has the main shortcoming that the number of estimated signals must be less than that of antenna array.For this fact,exploiting the non-circular characteristic of widely used communication signals,this paper present a virtual array based multiple signal classification( VA-MUSIC) algorithm.The new method combines the complex conjugate output signals with the original ones of the antenna array,so that the effective array aperture can be improved.Furthermore,by using the spatial smoothing technology,the DOAs of coherent signals can also be estimated.Computer simulations demonstrate that compared with the classical MUSIC,the proposed algorithm can provide much more accurate estimation,especially in the cases of lower signal-to-noise ratio( SNR) and smaller sampling number.