广义复合杂波模型参数估计方法  

Parameter estimation method for the generalized compound clutter model

作　　者：姜斌[1,2] 和湘[1] 任双桥[2] 郭桂蓉[2] 

机构地区：[1]通信指挥学院,湖北武汉430010 [2]国防科大四院空间信息技术研究所,湖南长沙410073

出　　处：《电路与系统学报》2009年第1期87-92,共6页Journal of Circuits and Systems

基　　金：国防预研资助项目(41303040203);国家自然科学基金资助(60402032)

摘　　要：广义复合杂波分布(Generalized Compound Probability Density Function,GC-PDF)模型是一种适用范围广泛的杂波分布模型,参数估计是研究该杂波模型的关键技术。本文首先建立了广义复合杂波模型,推导了其统计特性。在此基础上,研究了最小二乘参数估计方法。进而,以参数解耦和充分利用杂波序列信息为突破点,提出了一种新的杂波模型参数估计算法。该算法将一个四维非线性最优化问题转化为一个一维线性最优化问题,从而降低了计算量和所需样本数,提高了估计性能。最后,进行了仿真实验,验证了本文的结论。该算法对于高分辨雷达杂波分类和识别以及基于杂波背景的目标检测具有重要意义。Parameter estimation is the key problem to study the GC-PDF model, a clutter model which has a wide range of application. Firstly the GC-PDF model is set up, its statistical properties are deduced, and the least mean squares （LMS） parameter estimation method is studied on the model. Then, a new model parameter estimation method is put forward on the basis of parameter decoupling and full usage of the information contained in the clutter sequences. This new estimation method translates a four-dimension nonlinear optimization problem into a one-dimension linear optimization problem, hence reducing the calculation quantity and number of samples needed, and improving the estimation performance. Finally, the simulation is processed to verify the conclusion proposed in this paper. It is very important of this arithmetic to classify and identify high resolution radar clutters and to detect targets amid clutter background.

关 键 词：广义复合杂波分布 建模 参数估计 参数解藕 

分 类 号：TN955.3[电子电信—信号与信息处理]