基于细微特征提取的辐射源个体识别方法研究

发布时间：2018-10-22 12:20

【摘要】：基于信号细微特征分析的辐射源个体识别研究起源于非合作通信领域。所谓细微特征,指的是信号个体或设备个体由于发射机设备或者传输信道的影响,使得接收机接收到的信号所带有的能够作为个体身份标识的差异。区别于传统辐射源识别理论的是,传统的辐射源信号侦查识别的目的在于获取所传输的通信信息,而辐射源个体识别的目的是通过一定的信号处理过程,提取出隐藏在通信信息中的细微差异,从而识别、判断出对方辐射源的相关情报。如何选择有效的信号处理方法,实时、准确地分析、提取出这些细微差异特征是近些年来的研究热点。针对这一问题,本文深入研究了基于细微特征分析的辐射源信号以及设备个体识别的方法。论文的研究内容主要包括以下几个方面:建立了辐射源个体识别的系统模型,分析了细微特征产生的机理,研究了典型的辐射源细微特征提取方法,为后续细微特征分析新方法的研究提供了良好的理论基础。基于熵特征提取的辐射源识别方法。由于不同的信息熵能从不同的角度描述信号的差异性,本文提出了基于多维信息熵模型的识别模型,在此基础上研究了多维特征加权的方法。在信号识别环节,比较了基于欧氏距离、人工智能分类器以及所提出的特征加权方法的识别性能。研究了基于参数估计的辐射源识别方法。从细微特征分析的思路出发,提取出体现不同参数信号的个体特征,验证了所提取的特征在不稳定信噪比环境下的识别性能。然后利用特征非线性拟合的方法,来识别不同参数的线性调频信号个体。研究了基于振荡器非线性特征的辐射源识别方法。不同的通信设备由于自身器件非线性特征的差异,会使发送的信号含有设备的个体差异信息。提出了基于局部散布差异特征提取的设备非线性分析方法,并研究了其识别性能。
[Abstract]:Individual recognition of emitter based on signal fine feature analysis originates from the field of non-cooperative communication. The so-called fine characteristic refers to the signal individual or the device individual because of the transmitter equipment or the transmission channel influence, causes the receiver to receive the signal with the ability to act as the individual identification difference. Different from the traditional emitter recognition theory, the purpose of traditional emitter signal detection and recognition is to obtain the transmitted communication information, and the purpose of emitter individual identification is to pass a certain signal processing process. The subtle differences hidden in the communication information are extracted to identify and judge the relative information of the other side's emitter. How to select effective signal processing methods to analyze and extract these subtle features in real time and accurately is a hot topic in recent years. Aiming at this problem, the emitter signal based on fine feature analysis and the method of device individual identification are studied in this paper. The main contents of this paper are as follows: the system model of individual identification of emitter is established, the mechanism of fine feature generation is analyzed, and the typical methods of extracting subtle feature of radiation source are studied. It provides a good theoretical basis for the further study of the new method of fine feature analysis. Emitter recognition method based on Entropy feature extraction. Because different information entropy can describe the difference of signal from different angles, a recognition model based on multidimensional information entropy model is proposed in this paper. In signal recognition, the recognition performance based on Euclidean distance, artificial intelligence classifier and the proposed feature weighting method is compared. The emitter recognition method based on parameter estimation is studied. Based on the idea of fine feature analysis, the individual features of different parameter signals are extracted, and the recognition performance of the extracted features in unstable SNR environment is verified. Then the feature nonlinear fitting method is used to identify the individual of LFM signal with different parameters. The recognition method of emitter based on nonlinear characteristics of oscillator is studied. Because of the difference of the nonlinear characteristics of the devices, different communication devices will make the transmitted signals contain the individual difference information of the devices. A nonlinear analysis method based on local dispersion difference feature extraction is proposed and its recognition performance is studied.
【学位授予单位】：哈尔滨工程大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TN97

【参考文献】