基于循环平稳性的多制式宽带通信信号识别算法研究
发布时间：2018-12-29 18:55
【摘要】：通信信号调制方式是通信信号接收处理和承载信息提取的重要依据。在认知无线电、车载通信、空间电磁频谱监测、军用通信和电子对抗等应用领域中,都需要对未知信号进行感知和识别。接收机对接收到的一个或多个未知信号的处理除了信号检测、多域分析和参数估计等基本内容外,还必须通过调制识别判断信号调制类型,在此基础上加载对应的解调处理算法完成信号解调和传输信息的提取,进而实现其他任务。因此,调制识别是未知信息感知和获取中不可或缺的核心技术。本论文结合全新的图域数字信号处理(Digital Signal Processing on Graphs,DSP_G)理论,对宽带通信信号调制识别理论和方法进行研究。主要研究内容为基于循环平稳理论和DSP_G的通信信号调制方式识别理论和方法:1.根据通信信号循环谱的特性,研究并设计通信信号的图域映射与转换方法,图域映射的依据是循环谱矩阵的每个循环频率;2.根据图域映射与转换方法提取出的图,研究并设计相应的图域特征提取方法和图域分类器。首先提取训练信号和测试信号图集的邻接矩阵的特性,然后采用计算汉明距离实现图域分类器的设计;3.由于从理论上依据循环谱不能实现MPSK/MQAM信号的区分,因而,研究并设计了基于四阶循环累积量和DSP_G的MPSK/MQAM信号的识别方案,其中进行图域映射时需要将四阶循环累积量降维到二维矩阵,MQAM信号的类内识别是采用四阶累积量实现的。最后,本文对提出的基于循环谱和DSP_G的通信信号调制识别算法进行了仿真,同时和已存在的算法进行了对比,并详细分析了频偏和定时误差对提出算法性能的影响,详细推导分析了本文算法和已存在算法的计算复杂度。仿真表明提出的算法识别性能大大提升,虽然识别性能受频偏和定时误差的影响较大,但该算法的时间复杂度较低。在此基础上,对基于四阶循环累积量和DSP_G的MPSK/MQAM信号识别算法性能进行了仿真,仿真表明在MPSK/MQAM信号的识别中8PSK信号识别性能非常好。
[Abstract]:The modulation mode of communication signal is an important basis for receiving and processing communication signal and extracting carrying information. In the fields of cognitive radio, vehicular communication, space electromagnetic spectrum monitoring, military communication and electronic countermeasures, it is necessary to perceive and recognize unknown signals. In addition to signal detection, multi-domain analysis and parameter estimation, the receiver must judge the modulation type of the signal by modulation recognition, in addition to signal detection, multi-domain analysis and parameter estimation. On this basis, the corresponding demodulation processing algorithm is loaded to complete the signal demodulation and transmission information extraction, and then to achieve other tasks. Therefore, modulation recognition is an indispensable core technology in unknown information perception and acquisition. In this paper, the theory and method of modulation recognition for wideband communication signal are studied based on the new theory of digital signal processing (Digital Signal Processing on Graphs,DSP_G) in graphic domain. The main content of this paper is the theory and method of modulation recognition based on cyclic stationary theory and DSP_G: 1. According to the characteristics of cyclic spectrum of communication signal, the mapping and conversion method of communication signal is studied and designed. The mapping of graph domain is based on every cyclic frequency of cyclic spectrum matrix. 2. According to the graph extracted from map domain mapping and transformation method, the corresponding feature extraction method and image domain classifier are studied and designed. Firstly, the characteristics of the adjacency matrix of the training signal and the test signal graph set are extracted, and then the map domain classifier is designed by calculating the hamming distance. 3. Because the MPSK/MQAM signal can not be distinguished theoretically based on cyclic spectrum, the recognition scheme of MPSK/MQAM signal based on fourth-order cyclic cumulant and DSP_G is studied and designed. The fourth order cyclic cumulant is reduced to the two-dimensional matrix when mapping the graph domain, and the intra-class recognition of MQAM signal is realized by the fourth-order cumulant. Finally, the proposed modulation recognition algorithm based on cyclic spectrum and DSP_G is simulated and compared with the existing algorithms, and the effects of frequency offset and timing error on the performance of the proposed algorithm are analyzed in detail. The computational complexity of this algorithm and the existing algorithm is analyzed in detail. Simulation results show that the performance of the proposed algorithm is greatly improved. Although the recognition performance is greatly affected by frequency offset and timing error, the time complexity of the proposed algorithm is low. On this basis, the performance of MPSK/MQAM signal recognition algorithm based on fourth-order cyclic cumulant and DSP_G is simulated. The simulation results show that the recognition performance of 8PSK signal is very good in MPSK/MQAM signal recognition.
【学位授予单位】：电子科技大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TN911
本文编号：2395249
[Abstract]:The modulation mode of communication signal is an important basis for receiving and processing communication signal and extracting carrying information. In the fields of cognitive radio, vehicular communication, space electromagnetic spectrum monitoring, military communication and electronic countermeasures, it is necessary to perceive and recognize unknown signals. In addition to signal detection, multi-domain analysis and parameter estimation, the receiver must judge the modulation type of the signal by modulation recognition, in addition to signal detection, multi-domain analysis and parameter estimation. On this basis, the corresponding demodulation processing algorithm is loaded to complete the signal demodulation and transmission information extraction, and then to achieve other tasks. Therefore, modulation recognition is an indispensable core technology in unknown information perception and acquisition. In this paper, the theory and method of modulation recognition for wideband communication signal are studied based on the new theory of digital signal processing (Digital Signal Processing on Graphs,DSP_G) in graphic domain. The main content of this paper is the theory and method of modulation recognition based on cyclic stationary theory and DSP_G: 1. According to the characteristics of cyclic spectrum of communication signal, the mapping and conversion method of communication signal is studied and designed. The mapping of graph domain is based on every cyclic frequency of cyclic spectrum matrix. 2. According to the graph extracted from map domain mapping and transformation method, the corresponding feature extraction method and image domain classifier are studied and designed. Firstly, the characteristics of the adjacency matrix of the training signal and the test signal graph set are extracted, and then the map domain classifier is designed by calculating the hamming distance. 3. Because the MPSK/MQAM signal can not be distinguished theoretically based on cyclic spectrum, the recognition scheme of MPSK/MQAM signal based on fourth-order cyclic cumulant and DSP_G is studied and designed. The fourth order cyclic cumulant is reduced to the two-dimensional matrix when mapping the graph domain, and the intra-class recognition of MQAM signal is realized by the fourth-order cumulant. Finally, the proposed modulation recognition algorithm based on cyclic spectrum and DSP_G is simulated and compared with the existing algorithms, and the effects of frequency offset and timing error on the performance of the proposed algorithm are analyzed in detail. The computational complexity of this algorithm and the existing algorithm is analyzed in detail. Simulation results show that the performance of the proposed algorithm is greatly improved. Although the recognition performance is greatly affected by frequency offset and timing error, the time complexity of the proposed algorithm is low. On this basis, the performance of MPSK/MQAM signal recognition algorithm based on fourth-order cyclic cumulant and DSP_G is simulated. The simulation results show that the recognition performance of 8PSK signal is very good in MPSK/MQAM signal recognition.
【学位授予单位】：电子科技大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TN911
【参考文献】
相关期刊论文 前9条
1 Hamidreza Hosseinzadeh;;Tracking performance of large margin classifier in automatic modulation classification with a software radio environment[J];Journal of Systems Engineering and Electronics;2014年05期
2 宋鹏;张麟兮;张曼;唐勇;;基于时域平滑FFT累加算法的循环谱实现方法[J];电子设计工程;2013年11期
3 郑鹏;王泽众;刘锋;;循环谱时域平滑循环周期图法估计性能分析[J];计算机与数字工程;2011年11期
4 郭伟涛;王华力;甘泉;;基于TS201S的分段谱相关算法SSCA的实现[J];军事通信技术;2011年01期
5 孟玲玲;李静;;基于循环谱相关方法的MFSK信号识别[J];无线电通信技术;2010年01期
6 赵仕良;;MPSK信号的循环平稳特性分析[J];四川师范大学学报(自然科学版);2006年01期
7 张仔兵,李立萍,肖先赐;连续相位调制信号的二阶循环平稳特性分析[J];电子与信息学报;2005年11期
8 史建锋,朱良学;有限数据下循环谱的频域平滑对称式周期图法估计性能分析[J];数据采集与处理;2004年02期
9 韩国栋,蔡斌,邬江兴;调制分析与识别的谱相关方法[J];系统工程与电子技术;2001年03期
相关博士学位论文 前1条
1 高玉龙;基于循环谱的调制方式识别与高动态同步技术研究[D];哈尔滨工业大学;2007年
相关硕士学位论文 前1条
1 王睿;基于循环平稳性的数字信号调制识别与参数估计研究[D];电子科技大学;2012年
，本文编号：2395249
本文链接：https://www.wllwen.com/kejilunwen/xinxigongchenglunwen/2395249.html
