基于自适应子空间估计的DOA跟踪算法

运用特征子空间类高分辨方法的关键在于信号或噪声子空间的估计。实际上有些信号的统计特性通常随时间变化，为了得到参数的实时估计值，需要随时根据新的阵列接收数据对信号或噪声子空间进行更新?文中分析了一种自适应子空间估计算法，即MALASE(Maximum Likelihood Adaptive Subspaee Estimation)算法然后，把MALASE算法与最小范数(Mini—Norm)高分辨方位计算法相结合．并应用零点跟踪技术，提出了一种自适应Mini—Norm算法，可用于对时变的信号波达方向(DOA)进行跟踪估计。仿真结果验证了该算法具有较好的跟踪性能。

Algorithm for DOA tracking based on adaptive subspace estimation

The key problem of eigen-subspace high-resolution methods is estimation of signal or noise subspace. In practice, statistic characteristics of signals always change with time. To obtain real time estimates of the time varying signal parameters, it is necessary to constantly update the signal or noise subspace to adapt the newly sampled array output. In this paper, an algorithm termed MALASE(maximum likelihood adaptive subspace estimation)is presented to address the problem of adaptive estimation of the subspace of the data covariance matrix. It is based on optimization of a likelihood criterion. The parameters of the likelihood to be estimated are the expected eigenvectors and eigenvalues, obtained with a stochastic algorithm that requires little computation cost. Furthermore, the particular structure of the algorithm ensures the orthonormality of the estimated eigenvectors. Thus, with combination of the above-mentioned subspace tracking algorithm and a mini-norm high-resolution algorithm, and a zero-tracking technique, an adaptive mini-norm algorithm is proposed to track the time-varying DOAs (directions of arrival) of multiple targets. Computer simulation results are provided to demonstrate the effectiveness of the proposed algorithm.