基于统计特征值分解的雷达地杂波逆向仿真方法

Backward Simulation of Radar Ground Clutter Based on Factorizing Statistical Eigenvalue

任何平稳随机信号都可以用一些统计特征值来描述,得到这些统计特征值就可以采用数学的手段来反演出平稳随机信号。根据此思想,提出了一种不同于常规正向仿真法的雷达地杂波随机信号逆向仿真方法,该方法对地杂波协方差阵做cholesky因式分解来生成地杂波随机信号。这种方法适合于研究动目标检测等特殊后端处理算法,可避免正向仿真复杂耗时的计算过程。从仿真结果的统计特征值比较及对动目标检测算法的验证来看该方法是有效且方便的。

Any stationary stochastic signal can be characterized by some statistical eigenvalues, from which one can deduce the original stochastic signal by using mathematical methods. According to this idea, a new backward simulation approach which was absolutely different from the conventional forward one was proposed to simulate the radar clutter signal, based on cholesky-factorizing the clutter covariance matrix, This approach is suitable for researches about some special algorithm problems, e.g. moving target indicator. It can avoid the complex and time-consuming computing process involved in the conventional forward simulation approach. Compared with theoretical values and validated by moving-target-indicator processing, it is proved that this approach is feasible and convenient.