宽带无线电通信信号中的调制识别

发布时间：2019-01-29 05:17

【摘要】：在快速发展的无线通信技术领域,通信体制不断发展更新,通信信号的调制方式日益增加;以及各种通信之间相互干扰,通信环境日益复杂,因此,调制信号的自动调制识别技术在通信领域应用是非常广泛的,这其中包括在民用方面的应用和军事领域的应用,尤其是在非协作通信过程中,调制识别过程是必不可少的。此外,在卫星通信中,大多数采用MPSK(M-ary Phase Shift Keying)与MQAM(M-ary Quadrature Amplitude Modulation)等调制信号进行信息的传递,因此,这两类信号的类内调制识别是本文中的重点。本文主要对数字信号调制识别算法进行研究,对比各种调制识别算法的优缺点,结合现有的各种算法提出了能高效地对MPSK与MQAM信号进行类内识别的算法。本文主要分为如下四个部分:首先,对数字信号的调制识别过程进行介绍,目前调制识别的基本方式是基于模式识别的方法或者决策论的算法。其次,分析研究目前各种有关调制识别的基本方法,包括:基于时域特征参数提取的算法、基于最大似然估计的算法、基于幅度矩的算法、基于小波变换的算法,并对各种方法进行仿真。根据仿真结果进行比较,分析目前各方法的优缺点。然后,主要对MPSK信号的类内调制识别算法进行研究,在对各种算法的有效性进行对比之后,采用基于神经网络分类器的方法对信号进行调制识别。这种算法中,采用信号的高阶累积量作为识别的特征参数,分类器采用基于BP神经网络的分类器。仿真结果表明,此算法具有很好的抗噪声性能,在低信噪比下,对BPSK、QPSK、8PSK、16PSK等信号有很高的识别效率。最后,对MQAM信号的类内识别算法进行研究。主要采用了两种算法对MQAM信号进行类内识别:基于星座图聚类的算法、基于幅度值最大似然估计的方法。仿真结果表明,基于星座图聚类的算法在一定信噪比情况下,能有效识别4QAM、16QAM、32QAM、64QAM、128QAM、256QAM;而对矩形星座图的MQAM,采用最大似然估计的方法可以实现低信噪比下的调制识别。
[Abstract]:In the field of rapid development of wireless communication technology, the communication system is constantly developing and updating, and the modulation mode of communication signal is increasing day by day. And the communication environment is becoming more and more complex, so the automatic modulation recognition technology of modulated signals is widely used in the field of communication, including civil applications and military applications. Especially in the process of non-cooperative communication, modulation recognition is essential. In addition, in satellite communication, MPSK (M-ary Phase Shift Keying) and MQAM (M-ary Quadrature Amplitude Modulation) and other modulation signals are mostly used to transmit information. Therefore, the in-class modulation recognition of these two kinds of signals is the focus of this paper. In this paper, the modulation recognition algorithms of digital signals are studied, and the advantages and disadvantages of various modulation recognition algorithms are compared. Combined with the existing algorithms, an algorithm which can efficiently identify the MPSK and MQAM signals within the class is proposed. This paper is divided into four parts as follows: firstly, the modulation recognition process of digital signal is introduced. At present, the basic method of modulation recognition is based on pattern recognition or the algorithm of decision theory. Secondly, the basic methods of modulation recognition are analyzed and studied, including the algorithm based on time domain feature parameter extraction, the algorithm based on maximum likelihood estimation, the algorithm based on amplitude moment, and the algorithm based on wavelet transform. Various methods are simulated. According to the simulation results, the advantages and disadvantages of the current methods are analyzed. Then, this paper mainly studies the algorithm of intra-class modulation recognition of MPSK signal. After comparing the effectiveness of various algorithms, the method based on neural network classifier is used to identify the signal. In this algorithm, the high order cumulant of the signal is used as the feature parameter, and the classifier is based on the BP neural network. The simulation results show that the algorithm has good anti-noise performance and has high recognition efficiency for BPSK,QPSK,8PSK,16PSK signals under low SNR. Finally, the algorithm of MQAM signal recognition in class is studied. Two kinds of algorithms are mainly used to identify MQAM signals: constellation clustering algorithm and amplitude maximum likelihood estimation method. Simulation results show that the algorithm based on constellation clustering can effectively identify 4QAM16QAM-32QAM-64QAM-128QAM-256QAMunder a certain SNR. The maximum likelihood estimation (MLE) method for the MQAM, of the rectangular constellation can be used to realize the modulation recognition at low signal-to-noise ratio (SNR).
【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】：TN911.3

【参考文献】