时间平滑ML协方差矩阵估计算法及性能分析

以最少的快拍数得到最好性能的协方差矩阵估计是我们追求的目标,虽然这很难实现。由于对称阵列的协方差矩阵具有中央Hermitian性质,因而可以在协方差矩阵估计中加以利用。这种性质,体现在通过时间域平滑。分析给出了新算法的实现方法和概率密度函数,并通过MonteCarlo仿真,对算法的各种性能进行分析,表明该算法在相同快拍数时,比传统最大似然算法具有更小的误差性能。

Time Smoothing ML Estimation of Convariance Matrix and Performance Analysis

To get the best performance with the least snapshots is the ideal object when the covariance matrix is estimated,although it is not easy. However,the covariance matrix of a symmetrical array antenna has central Hermitian property. A new technique was developed to estimate the covariance matrix,which smoothes the estimated covariance matrix in time domain according to the central Hermitian property. The density function was proposed. The Monte Carlo simulations show that this algorithm has the less estimation bias than general maximum likelihood estimation even with the same snapshots.