基于模式的自制小波在EBPSK信号检测中的应用  

Pattern-based custom wavelet used for EBPSK signal detection

作　　者：张煜东[1,2] 朱庆[1] 陈贤卿[1] 吴乐南[1] 

机构地区：[1]东南大学信息科学与工程学院,南京210096 [2]哥伦比亚大学医学院,纽约10032

出　　处：《东南大学学报（自然科学版）》2011年第4期691-694,共4页Journal of Southeast University：Natural Science Edition

基　　金：国家自然科学基金资助项目(60872075);高等学校科技创新工程重大项目培育基金资助项目(706028);东南大学优秀博士学位论文基金资助项目(YBJJ0908)

摘　　要：为了改善基于小波的超窄带信号的检测率,提出了根据信号的模式构建小波的方法.以扩展二元相位键控的"1"信号为例,采用常数正交函数空间投影法构建了对应的unb1小波,并与经典的Haar,db系列、Symlet系列、Coiflet系列、Biorspline系列、逆Biorspline系列、离散美尔、高斯系列、墨西哥帽和Morlet小波进行比较.分析结果发现,所提出的unb1小波性能较好,能够检测出信号的位置和尺度,且在噪声强度大于信号强度的情况下亦可工作.因此,基于模式的自制小波用于超窄带信号检测是可行的.A novel custom wavelet based on signal patterns is proposed in order to improve the detection ratio of the UNB（ultra narrow band） signal.The ＂1＂ signal in extended binary phase shift keying（EBPSK） was taken as the example,and corresponding wavelet unb1 was constructed by orthconst method.Then,the proposed wavelet unb1 was compared with traditional wavelets including Haar,db series,Symlet series,Coiflet series,Biorsplines,reverse Biorsplines,discrete Meyer,Gaussian series,Mexican hat,and Morlet.The results indicate that the proposed unb1 is superior to traditional wavelets with respect to detecting both position and scale of the signal;besides,the method can work even if the magnitude of the noise is greater than that of the signal.In all,the pattern-based wavelet is feasible and effective in the application of UNB detection.

关 键 词：自制小波 超窄带 统一的二元相位调制和解调 信号检测 

分 类 号：TN911.23[电子电信—通信与信息系统]