一种基于累积法的K分布的参数估计方法  

Method for Estimating K-distribution Parameters Based on Accumulation Method

作　　者：高貂林 王新宏[1] 彭华[1] 

机构地区：[1]西北工业大学航海学院,西安710072

出　　处：《系统仿真学报》2014年第6期1297-1300,共4页Journal of System Simulation

摘　　要：提出了一种基于累积法的K分布参数估计方法,该方法利用变量替换将样本矩量和分布参数之间的关系转化为线性函数,然后通过累积法求解线性超定方程组获得K分布参数的估计。与最小二乘法求解该线性超定方程组相比,累积法可以对方程组的系数矩阵降维,使其变成二维的方阵,进而可以直接对该方阵求逆,降低了运算量。通过仿真对该方法进行了验证。A method for estimating the parameters of the K-distribution was proposed based on the accumulation method. This method used variable replacing to transform the relationship between the sample moment and distribution parameters into linear functions, and then the solution of linear overdetermined equations was sought by the accumulation method to estimate the parameters of K-distribution. Compared with the least square method, solving the linear overdetermined equations with the accumulation method could reduce the dimension of the coefficient matrix by making it into a two-dimensional square, then, could directly calculate the square matrix inverse, and reduced the computational complexity. This method was verified by simulation.

关 键 词：K分布 累积法 参数估计 仿真 

分 类 号：TN911.7[电子电信—通信与信息系统]