摘要
提出并采用信号的连续小波包分解的方法对BPSK、QPSK、8PSK数字调制信号进行模式识别。首先,根据3种调制信号的特点设置目标信号模板,然后将其分解到小波包子域形成子域模板并预先存储。执行识别任务时,将实际信号进行小波包分解,并在各小波包子域和子域模板作匹配相关,根据相关运算的结果完成识别。由于小波包分解的频带划分比小波分解的频带划分精细,所以识别的正确率更高。文中采用Monte Carlo法产生信号,采用Matlab进行仿真。仿真结果表明,采用小波包分解的识别正确率更高,尤其是在低信噪比环境下,该方法也能实现调制信号的精确识别。
In this study,automatic modulation recognition（AMR） of digitally phase shift keying signals,including BPSK,QPSK and 8PSK,by using continuous wavelet package transform（CWPT）,is presented. Firstly,according to the characteristic of each modulated signal,three templates of target signals are set,decomposed into wavelet package domain（WPD） to form sub-domain templates,and then stored in advance. While AMR is implemented,the real signal is decomposed into wavelet package domain（WPD） and then cross correlated with the subdomain templates. On the basis of the results of cross correlation,the modulation is identified. For the finer frequency division of WPD,the recognition rate is better than that by the general wavelet decomposition recognition.The signals are generated with Monte Carlo method. And the presented new method is simulated with Matlab. The outcomes of simulation show that the new method has a higher recognition rate,and particularly at low signal noise rate the method can precisely recognize the psk type of target signal.
出处
《电子科技》
2014年第9期124-127,共4页
Electronic Science and Technology
关键词
连续小波包变换
自动调制模式识别
相移键控
互相关
continuous wavelet package transform
automatic modulation recognition
phase shift keying
cross correlation
