实测数据的降秩空时自适应处理方法

在实际非均匀环境中，机载雷达数据的非均匀性会恶化杂波协方差矩阵的估计，严重地降低了空时自适应处理器的性能。本文首先用非均匀检测器在非均匀环境下选择辅助距离门，来估计样本协方差矩阵。而后讨论了基于广义旁瓣相消器结构上的两种降秩空时自适应处理方法：主分量法和互谱法，这两种方法都是利用杂波和干扰的低秩属性，用依赖于采样数据的转换矩阵来构造空时自适应滤波器。最后对从Mountain Top计划测得的数据，应用上述两种降秩方法对其进行了分析。仿真实验表明了这两种方法的可行性，减少了计算量，并提高了有效可测收敛性，可有效地提高空时自适应处理器的性能。

Reduced-rank Space-Time Adaptive Processing to Measure Data

In practical non-homogeneous environment, due to the non - homogeneity of airborne radar data, the estimation of the covariance matrix from non-homogeneous training data resulted in severely undernulled clutter, and space-time adaptive processor performance suffered. This paper firstly introduced the non-homogeneity detection (NHD), to select the secondary range cells and estimate the correlation matrix in the non-homogeneous environment. And then discussed reduced-rank space-time adaptive processing (STAP) based on the generic sidelobe canceller (GSC): eigen-based principal component and cross-spectral metric. Those two approaches all take advantage of the low rank nature of clutter and jamming observations, and the reduced-dimension transformation applied to the data are necessarily data dependent. Lastly using the measure data from the Mountain Top program conducted the two reduced-rand STAP technique. The measure data analysis showed that those two approaches reduced computational burden and improved convergence measure of effectiveness. Significant performance improvement can be achieved.