图2.11a比较了指数函数、四自由度chisquare函数、二自由度非中心chi-square函数、威布尔函数和对数正态密度函数，这些函数的RCS方差均为0.5。

Figure 2.11a compares the exponential, fourth-degreechisquare, second-degree non-central chi-square, Weibull, and log-normaldensity functions when all have an RCS variance of 0.5.

Figure 2.11. 雷达截面概率密度函数的五种模型比较：（a）线性尺度，（b）对数尺度Comparison of five models forthe probability density function of radar cross section: (a) linear scale, (b)log scale. See text for additional details.

因此，指数分布的平均值为0.5。

The exponentialdistribution then necessarily has a mean of 0.5.

四自由度chi-square函数的平均值为1.0，其它密度函数的平均值也应当选择为1.0。

The fourth-degreechi-square necessarily has a mean of 1.0, and the parameters of the remainingdensity functions have been chosen to give them a mean of 1.0 as well.

图2.11b以半对数刻度重复描绘了相同的数据，这样PDF的拖尾特征就更加明显了。

Figure 2.11b repeatsthe same data on a semilogarithmic scale so that the behavior of the PDF"tails" is more evident.

注意，威布尔和二自由度非中心chi-square分布对于这种参数的选择是非常相似的。

Note that the Weibulland second-degree non-central chi-square distributions are very similar forthis choice of parameters.

chi-square分布也是相似的，但该分布的拖尾相对较小。

The chi-square isalso similar, but has a somewhat less extensive tail to the distribution.

对数正态分布既有最窄的峰值范围，也有最长的拖尾。

The log-normal hasboth the narrowest peak and the longest tail of any of the distributions shownfor this choice of parameters.

与所有其它的分布函数不同，指数分布在平均RCS附近没有明显的峰值。

Unlike all of theothers, the exponential does not have a distinct peak near the mean RCS.

其它分布都有一个明显的峰值，这使得它们适合于描述一个或几个主要散射体的RCS特征。

Each of the othersdoes have a distinct peak, making them suitable for distributions with one or afew dominant scatterers.

——本文译自Mark A. Richards所著的《Fundamentals of Radar Signal Processing（Second edition）》

更多精彩文章请关注微信号：