基于矩阵拟合的InSAS干涉相位估计方法

An interferometric phase estimation method for In SAS based on matrix fitting

针对干涉合成孔径声呐相位估计问题,提出一种基于矩阵拟合的干涉相位估计方法。该方法通过构造联合观测数据矢量获取邻近像素的相干信息,通过对观测数据矢量对应的联合协方差矩阵进行拟合实现干涉相位的准确估计,优点是能够避免子空间分解类方法中存在的信号子空间扩散问题。所提方法通过给出干涉相位闭式解的形式,解决了角度搜索导致的大运算量问题。仿真和实测数据的处理结果证明了方法的有效性。

An interferometric phase estimation method based on matrix fitting is proposed for synthetic aperture sonar interferometry （InSAS） phase estimation. In this method, the joint data vector is formulated for making use of inter- ferometric information embedded in the neighboring pixels and the matrix fitting technique is used for joint covari- ance matrix corresponding to observation data vector, which makes it possible to accurately estimate the interfero- metric phase embedded in joint covariance matrix. The benefit is to avoid the signal subspace problem that exists in the subspace decomposition method. In this paper, a closed-form solution of interferometric phase is given as an al- ternate to the angle-scanning search based method since it is time-consuming. Simulation results and processing of real data demonstrated the validation of the method.