部分均匀环境中训练样本不足时的贝叶斯子空间检测器
Bayesian subspace detector with limited training data in partially homogeneous environments
周喆 1刘维建 2吴云韬 1郑岱堃 2巩朋成1
作者信息
- 1. 武汉工程大学计算机科学与工程学院,湖北武汉 430205;武汉工程大学智能机器人湖北省重点实验室,湖北武汉 430205
- 2. 空军预警学院,湖北武汉 430019
- 折叠
摘要
为了解决部分均匀环境中训练数据不足时的子空间信号检测难题,采用贝叶斯理论,将噪声协方差矩阵建模为逆威沙特分布,并采用广义似然比准则(generalized likelihood ratio test,GLRT)、Rao准则和 Wald准则设计自适应检测器,结果表明3种准则得到相同的结果.基于仿真及实测数据验证了所提检测器的有效性,并得出了影响检测性能的关键物理量.
Abstract
To solve the problem of subspace signal detection in partially homogeneous envi-ronment with limited training data,the Bayesian theory was adopted by modelling the un-known covariance matrix as an inverse Wishart distribution.Then,an adaptive detector was designed according to the generalized likelihood ratio test(GLRT),Rao test and Wald test.Numerical examples of simulations and real data have demonstrated that the proposed detec-tor can provide better detection performance than those existing detectors.The key physical quantities that affect the detection performance were also obtained.
关键词
目标检测/GLRT/Rao准则/Wald准则Key words
target detection/GLRT/Rao test/Wald test引用本文复制引用
基金项目
国家自然科学基金(62071482)
国家自然科学基金(62071172)
湖北省自然科学基金(2023AFA035)
湖北省重点研发计划(2022BAA052)
湖北三峡实验室开放基金(SC215001)
湖北省教育厅重点项目(D20221504)
湖北省教育厅科学技术研究项目(B2022062)
出版年
2024