摘要
Pareto分布是一种重要的非高斯分布,被证明能够有效描述高分辨率主动声纳的混响统计特性。文章分析了有序统计选小( Ordered Statistic with Smallest Option, OSSO )和有序统计选大( Ordered Statistic with Greatest Option, OSGO )两种恒虚警( Constant Fales Alarm Rate, CFAR )检测器在Pareto分布混响背景下的性能。在尺度参数已知情况下,证明了OSSO-CFAR和OSGO-CFAR对形状参数具有恒虚警的特性。进一步分析了两种检测器在均匀Pareto混响背景、多目标干扰及混响边缘情况下的性能,并与有序统计( Ordered Statistic, OS ) CFAR进行了对比。结果表明,在均匀混响背景下,OSGO-CFAR的检测性能与OS-CFAR相近,在混响边缘情况下具有最好的虚警控制能力;而对于多目标干扰情况,OSSO-CFAR比其他两种检测器的检测性能更优。
The Pareto distribution was an important non-Gaussian distribution. It has been proven to effectively describe the reverberation statistical properties of high-resolution active sonar. The performance of the ordered statistic smallest option (OSSO) and the ordered statistic greatest option (OSGO) of constant fales alarm rate (CFAR) detectors in Pareto distribution reverberation background was analyzed in this paper. Assuming the scale parameter was known, it was proved that OSSO-CFAR and OSGO-CFAR detectors also have constant false alarm performance on shape parameter. And the performance of two detectors was analyzed against Pareto homogeneous reverberation background, multi-target interference and edge-reverberation environments, and it was compared with the ordered statistic (OS) CFAR detector. The results show that the detection performance of the OSGO-CFAR detector is very similar to OS-CFAR detector in homogeneous background and it has the best false alarm control ability in the case of edge-reverberation. For multi-target interference environments, OSSO-CFAR detector performs better than the other two detectors.
作者
魏嘉
徐达
闫晟
郝程鹏
Wei Jia;Xu Da;Yan Sheng;Hao Chengpeng(Institute of Acoustics, Chinese Academy of Sciences, Beijing 100190, China;University of Chinese Academy of Sciences, Beijing 100049, China)
出处
《信号处理》
CSCD
北大核心
2019年第9期1599-1606,共8页
Journal of Signal Processing
基金
国家自然科学基金资助项目(61571434)
关键词
PARETO分布
检测
均匀背景
干扰目标
混响边缘
Pareto distribution
detection
homogeneous background
interfering targets
edge-reverberation
