一维GTD散射中心模型参数估计的改进MUSIC算法

针对利用多重信号分类（MUSIC）算法估计一维几何绕射理论（GTD）的散射中心模型时噪声鲁棒性较差、参数精度不高这一问题，提出一类改进的MUSIC算法。首先，构建原始回波数据的共轭矩阵，有效提高了原始回波数据的利用率；其次，将原始回波数据的协方差矩阵、共轭数据的协方差矩阵叠加取平均，得到一个新的总协方差矩阵；最后，对矩阵作2次方、4次方等偶次方处理，得到另一矩阵，以达到增大信号特征值与噪声特征值之间差距的作用，等效为增大了信噪比。仿真结果表明:所提改进算法的参数估计性能及噪声鲁棒性均要优于经典MUSIC算法。 

Parameter estimation of 1D GTD scattering center model based on an improved MUSIC algorithm

The noise robustness and parameter accuracy are poor when the classical Multiple Signal Classification (MUSIC) algorithm is used to estimate parameters of the one-dimensional Geometric Theory of Diffraction (GTD) scattering center model. To solve this problem, a series of improved MUSIC algorithms are proposed in this paper. Firstly, the improved algorithms construct the conjugate matrixof the original back-scattered data, which utilizes the information of the original data more effectively. Secondly, by averaging the covariance matrix of the original scattering data and its conjugated data, a novel total covariance matrix can be obtained. Finally, quadratic, quartic and other even power are performed on the matrix to obtain another matrix, and thus it can broaden the differences between the eigenvalues of noises and signals, which is equivalent to increasing the signal-to-noise ratio. Simulation results show that the parameter estimation performance and noise robustness are better than those of the classical MUSIC algorithm.