短波跳频信号盲检测技术研究

发布时间：2019-05-18 15:34

【摘要】：跳频通信具有抗干扰能力强、截获概率低、抗衰落能力强和多址组网能力等优势,其在短波通信中的应用日益广泛,针对跳频信号的研究已成为信号监测中的重要内容。作为非协作第三方,快速有效地从复杂电磁环境中检测出跳频信号,是对其进行后续处理的前提,也是跳频信号监测的难点。对此本文主要研究了短波跳频信号的盲检测技术,主要工作和创新如下:1.研究了短波宽带接收的预处理问题。首先介绍了短波宽带接收的复杂电磁环境,并分析了带内不平坦噪声基底对信号检测的影响。然后分析了基于低通滤波和形态学滤波的两种噪声基底估计算法,并采用高斯型结构元素对形态学估计算法进行改进,使得白化处理后的背景噪声近似为白噪声,为后续的短波跳频信号盲检测奠定了基础。2.研究了强干扰条件下跳频信号的存在性判定问题。在信道化检测算法的基础上,根据不同信号之间频谱的概率分布不同,给出了基于时频图单频分量谱熵和加权标准差的两种跳频信号存在性检测算法。首先给出了单频分量的谱熵和加权标准差两个概率分布特征的统计量,并对不同信号类型的谱熵和加权标准差进行分析。然后根据其特性给出了谱熵法和加权标准差法的跳频信号存在性检测算法。仿真结果表明,该算法无需先验信息,能够有效判定是否存在跳频信号,为后续的进一步检测提供依据。3.研究了基于信号灰度时频图的跳频信号盲检测问题。首先根据跳频hop在时频图中的邻域分布特征及其空间位置信息,引入对称共生矩阵阈值分割算法对灰度时频图处理,给出了一种基于改进对称共生矩阵的跳频信号盲检测算法。然后根据跳频hop在时频图中的灰度形态特征,分别对灰度时频图的频率分量和时间分量进行灰度形态学滤波,给出了一种基于二次形态学滤波的跳频信号盲检测算法。仿真结果表明所提算法能够有效降低噪声和干扰信号的影响,从灰度时频图中提取完整的跳频hop,且无需对时频图二值化处理,算法简单易于工程实现。4.研究了基于信号向频图的跳频信号盲检测问题。针对接收带宽内同时存在多个跳频信号的情况,引入信号示向信息对现有检测算法进行改进。首先给出了向频图的定义,并分析了其用于跳频信号检测的可行性。然后根据信号示向在向频图中的分布,分别进行信号示向检测、二值化分割向频图和HMT提取特征等处理,给出了一种基于示向分析的检测算法。最后根据前两章的研究结论,联合时频图对向频图进行预处理,有效地降低了噪声示向的影响。仿真结果表明该算法能够有效地分离信号,省去了复杂的去干扰步骤,且联合时频图预处理,提高了信号示向的检测概率并降低了虚警。
[Abstract]:Frequency hopping communication has many advantages, such as strong anti-interference ability, low probability of interception, strong anti-fading ability and multi-access networking ability, and its application in shortwave communication is becoming more and more extensive. The research on frequency hopping signal has become an important part of signal monitoring. As a non-cooperative third party, fast and effective detection of frequency hopping signal from complex electromagnetic environment is the premise of subsequent processing, and it is also the difficulty of frequency hopping signal monitoring. In this paper, the blind detection technology of shortwave frequency hopping signal is studied. The main work and innovation are as follows: 1. The preprocessing problem of shortwave broadband reception is studied. Firstly, the complex electromagnetic environment of shortwave broadband reception is introduced, and the influence of in-band uneven noise substrate on signal detection is analyzed. Then two kinds of noise base estimation algorithms based on low-pass filtering and morphological filtering are analyzed, and the Gaussian structural elements are used to improve the morphological estimation algorithm, so that the background noise after whitening is approximately white noise. It lays a foundation for the subsequent blind detection of shortwave frequency hopping signals. 2. In this paper, the existence of frequency hopping signals under strong interference is studied. On the basis of channelized detection algorithm, according to the different probability distribution of spectrum among different signals, two kinds of frequency hopping signal existence detection algorithms based on spectral entropy and weighted standard deviation of single frequency component of time-frequency graph are presented. Firstly, the statistics of spectral entropy and weighted standard deviation of single frequency component are given, and the spectral entropy and weighted standard deviation of different signal types are analyzed. Then, according to its characteristics, the existence detection algorithm of frequency hopping signal based on spectral entropy method and weighted standard deviation method is given. The simulation results show that the algorithm does not need prior information and can effectively determine whether there is a frequency hopping signal or not, which provides a basis for further detection. The blind detection of frequency hopping signals based on gray time frequency graph is studied. Firstly, according to the neighborhood distribution characteristics and spatial position information of frequency hopping hop in time frequency graph, the threshold segmentation algorithm of symmetric symbiosis matrix is introduced to process gray time frequency graph, and a blind detection algorithm of frequency hopping signal based on improved symmetric symbiosis matrix is proposed. Then, according to the gray morphological characteristics of frequency hopping hop in time frequency map, the frequency component and time component of gray time frequency graph are filtered by gray morphology, and a blind detection algorithm of frequency hopping signal based on quadratic morphological filtering is proposed. The simulation results show that the proposed algorithm can effectively reduce the influence of noise and interference signals, extract the complete frequency hopping hop, from the gray time frequency diagram without binarization of the time frequency graph, and the algorithm is simple and easy to be implemented in engineering. 4. The blind detection of frequency hopping signals based on signal directed frequency graph is studied. In view of the fact that there are multiple frequency hopping signals in the receiving bandwidth at the same time, the signal direction information is introduced to improve the existing detection algorithms. Firstly, the definition of directed frequency diagram is given, and its feasibility for frequency hopping signal detection is analyzed. Then, according to the distribution of signal direction in directed frequency diagram, signal direction detection, binarization segmentation direction frequency graph and HMT feature extraction are carried out respectively, and a detection algorithm based on direction analysis is proposed. Finally, according to the conclusions of the first two chapters, the directional frequency diagram is preprocessed by combining time-frequency diagram, which effectively reduces the influence of noise direction. The simulation results show that the algorithm can effectively separate the signal, save the complex de-interference steps, and combine time-frequency graph preprocessing, which improves the detection probability of signal direction and reduces the false alarm.
【学位授予单位】：解放军信息工程大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TN914.41

【参考文献】