小快拍高分辨目标方位估计算法 GMUSIC的性能分析*

针对水下运动阵列在运动过程中进行方位估计时存在快拍不足的问题,研究了基于随机矩阵理论的MUSIC改进算法GMUSIC,该方法通过Stieltjes变换建立起统计协方差矩阵真实特征值、特征向量与样本协方差矩阵之间在逼近域中的关联,以修正样本协方差特征分解的结果,进而实现小快拍方位估计。仿真与试验表明:GMUSIC算法可以更好地分辨相邻目标,且需要的快拍数较MUSIC算法要少;在低信噪比情况下,GMUSIC算法方位估计均方根误差远小于MUSIC算法,估计成功概率远大于MUSIC算法。因此,GMUSIC算法适用于解决水声目标的小快拍方位估计问题。

Performance analysis of GMUSIC algorithm for high resolution target bearing estimation in small snapshots

When bearing estimation is done with underwater moving array, in order to solve the issue of insufficient snapshots, the performance of an improved MUSIC algorithm GMUSIC based on Random Matrix Theory is analyzed . The GMUSIC method uses Stieltjes transformation to establish the correlation between the true eigenvalues of the statistical covariance matrix and the sample covariance matrix in the approximation domain, to correct the result of the smaple covariance eigendecomposition. Simulations and experiments show that the GMUSIC algorithm can better distinguish adjacent targets and require fewer snapshots than the MUSIC algorithm. In the case of low SNR, the root mean square error of the GMUSIC algorithm is much smaller than that of the MUSIC algorithm. The probability of success is much greater than the MUSIC algorithm. Therefore, the GMUSIC algorithm is suitable for solving the problem of position estimation of small snapshots of underwater acoustic targets.