论文部分内容阅读
本文介绍了一类新的恒虚警率(CFAR)处理器。Ll-CFAR通过线性地从参考装置上滤去分级样本来进行噪声功率的估测,这种组合的权数不仅取决于等级,还依赖于样本和检测单元的相对接近度。从Ll-CFAR的一大类中,可以选择出(1)能有效地探测出假目标的;(2)能在杂波边界存在的情况下有效地控制虚警的;(3)适用上述两种非均相性情况的具有韧性的方案。当按这些方案进行设计时,其实际工作量并不比单纯的有序统计CFAR(OS-CFAR)大。本文在对Ll-CFAR的随机训练进行讨论之后,在所遇到的最普遍的环境条件下,对其性能进行了充分的评估,并且将其性能与传统的CFAR技术的性能作了比较。
This article describes a new class of CFAR processors. Ll-CFAR performs noise power estimation by filtering out the graded samples linearly from the reference device. The weight of this combination depends not only on the level but also on the relative proximity of the sample and the detection unit. From a broad class of Ll-CFARs, one can choose (1) to detect false targets effectively; (2) to effectively control false alarms in the presence of clutter boundaries; (3) Toughness scenario with heterogeneous conditions. When designed according to these scenarios, the actual workload is no larger than the simple ordered statistical CFAR (OS-CFAR). After discussing the random training of Ll-CFAR, this paper fully evaluates its performance under the most common environmental conditions and compares its performance with the performance of traditional CFAR technology.