首页 | 本学科首页   官方微博 | 高级检索  
     

广义相关检测器中噪声增强条件的研究
引用本文:张静,许丽艳,段法兵.广义相关检测器中噪声增强条件的研究[J].青岛大学学报(工程技术版),2013(4):53-57.
作者姓名:张静  许丽艳  段法兵
作者单位:青岛大学复杂性科学研究所,山东青岛266071
基金项目:山东省自然科学基金项目资助(ZR2010FM006)
摘    要:为了检测在噪声存在的情况下弱信号是否出现,本文对噪声增强广义相关检测器性能的条件进行了研究.通过观测数据,给定虚警概率,证明了一类广义相关检测器中随机共振现象的存在性,并给出了噪声增强广义相关检测器性能的必要条件及求解最优尺度噪声强度的解析方程.研究结果表明,当观测数据足够大时,广义相关检测器的检测概率是检测功效的单调函数;噪声与检测器结构参数都可以视为一种增强检测器性能的调谐因素.该方法作为一种新的信号检测技术,对非线性信号处理的发展具有重要意义.

关 键 词:广义相关检测器  检测功效  尺度噪声  随机共振

Study of Noise-Enhanced Conditions for the Generalized Correlation Detector
ZHANG Jing,XU Li-yan,DUAN Fa-bing.Study of Noise-Enhanced Conditions for the Generalized Correlation Detector[J].Journal of Qingdao University(Engineering & Technology Edition),2013(4):53-57.
Authors:ZHANG Jing  XU Li-yan  DUAN Fa-bing
Affiliation:(Institute of Complexity Science, Qingdao University, Qingdao 266071, China)
Abstract:In order to detect the weak signal in additive white noise, this paper studies the noise-enhanced detection condition for a generalized correlation detector. Given the false probability, the occurrence of sto- chastic resonance effect is demonstrated in the considered generalized correlation detector, and the positive derivative of the efficacy with respect to the noise level is the necessary condition of the detector perform- ance enhancement. The optimal scaled-noise level that maximizes the detector efficacy can be analytically solved. The obtained results show that, with a large number of observations, the detection probability of a generalized correlation detector is a monotonically function of the detection efficacy for detecting the weak signal in additive white noise. Noise, as the same as the detector parameters, is a tunable factor of enhan- cing the detector performance. The noise-enhancement method becomes a novel signal detection technique, which is meaningful for the field of nonlinear signal processing.
Keywords:generalized correlation detector  detection efficacy  scaled-noise  stochastic resonance
本文献已被 维普 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号