基于扩展孔径ESPRIT算法的高精度无模糊二维DOA估计

针对扩展阵列孔径会产生估计模糊的问题，本文提出一种基于双平行阵扩展孔径ESPRIT算法。算法首先设置双平行阵传感器之间距离远大于半波长以获得阵列扩展孔径；然后将传感器按照平行于x轴与平行于y轴分为2个子阵；对平行于x轴的子阵采用ESPRIT法，计算出低精度但不模糊的y轴方向余弦估计和高精度但模糊的x轴方向余弦估计，同样由平行于y轴的子阵算出低精度但不模糊的x轴方向余弦估计和高精度但模糊的y轴方向余弦估计； 最后将低精度但不模糊的方向余弦估计作为参考值来解由于扩展孔径带来的模糊问题，并得到最终的波达方向估计值。本文提出的算法无需二维搜索，是一种低运算量高精度的算法。此外本文还针对特征值相等时的角度配对失效现象，提出一种基于子空间正交的新配对方法。计算机仿真实验验证了本文算法的性能。

High Accurate and Unambiguous 2D Direction Finding Estimation Based on Extended Aperture ESPRIT Algorithm

A new high accurate azimuth and elevation angle estimation algorithm using ESPRIT algorithm is developed to avoid the ambiguous of extended aperture.Firstly the two-parallel-shape array geometry,which consists of sensors spaced much farther apart than a half-wavelength,is exploited to achieve array aperture extension,and the sensors are divided into 2 sub-arrays.The sub-array parallel with the x-axis is employed to extract low accurate but unambiguous y-axis direction cosines and high accurate but cyclically ambiguous x-axis direction cosines.Then,the sub-array parallel with the y-axis is employed to extract unambiguous x-axis direction cosines and cyclically ambiguous y-axis direction cosines.Finally,the low accurate unambiguous direction cosine estimates are used to resolve the cyclically ambiguous direction cosine estimates to obtain azimuth-elevation angle estimates.The proposed algorithm requires no 2D iterative searching,and thereby,offering high estimation precision with low computational complexity.Furthermore,a new pairing method based on the orthometric subspaces is proposed for the equivalent eigenvalues case.The simulation results indicate the performance of proposed algorithms.