在现代化战争中，无论是指挥控制还是情报探测都必须通过通信系统实现信息的传输和交换。因此在复杂电磁环境下，对非合作方通信系统精准实施干扰并且对干扰效果进行评估非常重要。通信系统的受干扰程度决定了信息交换的质量，进而影响着战争的胜负。在非合作通信场景下，若能够基于侦察截获的目标通信数据，进行通信信号类型识别，根据识别结果实施针对性的干扰，再对其通信系统的受扰情况进行分析评估，这对己方干扰方案的制定和优化有着重要的意义。本文根据新形式下电子战的发展要求，围绕着非合作场景下通信信号类型识别技术及通信干扰效果评估技术展开研究，为战时通信博弈提供技术支持和理论依据。论文重点研究的内容如下：

首先，论文设计出一种非合作通信干扰效果评估方法，该方法首先利用智能处理方法对非合作通信信号进行识别，确定通信调制类型，再依据论文推导的多种干扰样式下通信信号的误码曲线，结合信干比等参数，可给出非合作通信干扰效果的定量分析。论文研究了常见通信干扰信号的产生原理，仿真生成了单音、多音、窄带、宽带、脉冲和瞄准式这6种类型干扰信号，分析了不同干扰信号时频域特征。

然后，针对非合作场景下如何快速识别敌方通信信号，提出了一种卷积神经网络混合输入智能信号识别方法。考虑了信号在时域和频域两种不同的表达形式，以通信信号的频谱图和原始的IQ数据作为两个子卷积网络的输入，利用卷积神经网络从输入数据中自动提取特征。然后对两个子网络提取的特征进行融合送入最后的分类层来完成信号的识别任务。通过仿真产生6种常见的通信信号，在不同的信噪比下产生大量的样本分别组成训练集和测试集。网络训练结果表明，混合输入的识别方法的性能要远远优于只使用单独输入的识别方法，在信噪比大于10dB时，6种类型信号识别准确率都能达到90%以上。

最后，本文从BPSK、2FSK、QPSK、16QAM、32QAM和64QAM调制信号产生机理出发，结合通信信号受扰的成因，理论推导给出了BPSK、2FSK、QPSK、 16QAM、32QAM和64QAM调制信号在受到单音、多音、窄带和宽带干扰后的误码率表达式；论文仿真了不同配置下的性能曲线，所推导的不同调制类型遭遇不同干扰样式的受扰误码性能曲线，理论值与蒙特卡洛仿真值高度一致，验证了论文所做干扰效果评估理论的正确性；论文提供的干扰性能评估方法还可推广到更多的干扰样式和调制类型，为非合作通信干扰性能定量评估提供可解释的依据，该评估结果可用于驱动干扰资源的配置。此外，将上述信号源生成、通信信号类型识别及干扰效果评估算法模块进行系统集成，基于Qt平台开发软件界面并进行演示验证。

综上，在未来通信博弈对抗中，尤其是非合作场景下，首先要做到“先敌发现”即利用智能化手段精准快速识别非合作方通信信号样式，从而采取针对性的干扰措施对非合作方通信系统进行破坏；然后还要做到“预先评估”即利用我方先验信息及通信信号受扰性能评估模型对非合作方通信系统在可能受到我方干扰后的通信性能进行事先评估，做到“心中有数”；最后结合评估结果和实时监测数据，分析非合作方通信行为的变化情况，进而对我方的干扰手段、通信参数、网络模型参数等进行相应调整，以保证我方对于非合作方通信系统的干扰优势。

In modern warfare, information transmission and exchange must be realized through communication system, whether it is command and control or intelligence detection. Therefore, in complex electromagnetic environment, it is very important to accurately implement jamming to non-cooperative communication systems and evaluate the jamming effect. The degree of interference of communication system determines the quality of information exchange and thus affects the outcome of war. In the non-cooperative communication scenario, if the communication signal type can be identified based on the target communication data intercepted by reconnaissance, the targeted interference can be implemented according to the identification results, and then the interference situation of the communication system can be analyzed and evaluated, which is of great significance for the formulation and optimization of the interference scheme of our own side. According to the development requirements of electronic warfare in the new form, this paper studies the type identification technology of communication signal and the evaluation technology of communication interference effect in the non-cooperative scenario, providing technical support and theoretical basis for the wartime communication game. The thesis focuses on the following contents:

 

Firstly, a non-cooperative communication interference effect evaluation method is designed in this paper. The method firstly uses intelligent processing method to identify non-cooperative communication signals and determine the communication modulation type. Then, according to the error curve of communication signals under various interference styles derived in this paper, combined with the signal-to-dry ratio and other parameters, the interference effect of non-cooperative communication can be quantitatively analyzed. In this paper, the generation principle of common communication interference signals is studied. Six types of interference signals are simulated, namely single-tone, multi-tone, narrowband, broadband, pulse and aiming, and the time-frequency domain characteristics of different interference signals are analyzed.

 

Then, a convolutional neural network hybrid input intelligent signal recognition method is proposed to quickly identify enemy communication signals in non-cooperative scenarios. Two different expression forms of signal in time domain and frequency domain are considered. The spectrum diagram of communication signal and the original IQ data are used as the input of two sub-convolutional networks, and the features are automatically extracted from the input data by convolutional neural network. Then the features extracted from the two subnetworks are fused into the final classification layer to complete the task of signal recognition. Through simulation, 6 kinds of common communication signals are generated, and a large number of samples are generated under different signal-to-noise ratios to compose the training set and the test set respectively. The results of network training show that the performance of the recognition method with mixed input is much better than that with single input. When the signal-to-noise ratio is greater than 10dB, the recognition accuracy of six types of signals can reach more than 90%.

 

Finally, this paper starts from the generation mechanism of BPSK, 2FSK, QPSK, 16QAM, 32QAM and 64QAM modulated signals, and combines the cause of interference of communication signals. The bit error rate expressions of BPSK, 2FSK, QPSK, 16QAM, 32QAM and 64QAM modulated signals subjected to single-tone, multitone, narrowband and wideband interference are derived theoretically. In this paper, the performance curves of different configurations are simulated. The derived performance curves of different modulation types encounter different interference styles, and the theoretical values are highly consistent with the Monte Carlo simulation values, which verifies the correctness of the interference evaluation theory proposed in this paper. The evaluation method presented in this paper can also be extended to more interference styles and modulation types, providing an interpretable basis for the quantitative evaluation of non-cooperative communication interference performance, and the evaluation results can be used to drive the allocation of interference resources. In addition, the above signal source generation, communication signal type recognition and interference effect evaluation algorithm modules are integrated, and the software interface is developed based on Qt platform and demonstrated and verified.

 

To sum up, in the future communication game confrontation, especially in the non-cooperation scenario, the first thing to be done is to "discover the enemy before the enemy", that is, to use intelligent means to accurately and quickly identify the communication signal pattern of non-partner, so as to take targeted interference measures to destroy the communication system of non-partner. Then it is necessary to "pre-evaluate", that is, to use our prior information and communication signal disturbance performance evaluation model to pre-evaluate the communication performance of non-partner's communication system which may be interfered by us, so as to "know clearly". Finally, combined with the evaluation results and real-time monitoring data, the change of communication behavior of non-partner is analyzed, and then the interference means, communication parameters and network model parameters of our side are adjusted accordingly, so as to ensure the interference advantage of our side to the communication system of non-partner.

[1] LI X B WANG W P, LIN M, et al. The research framework, progress, and key directions of system-of-systems contribution ratio evaluation[J]. Systems Engineering-Theory & Practice, 2019, 39(6): 1623-1634.
[2] 李正东. 典型信号调制类型识别方法研究[D]. 四川:电子科技大学,2013.
[3] 李小捷,许录平,陈佳. 基于联合参数估计的最大似然调制识别算法[J]. 仪器仪表学报,2008,29(12):2509-2514.
[4] WEI W, MENDEL JM. Maximum-likelihood classification for digital amplitude-phase modulations[J]. IEEE Transactions on Communications,2000,48(2):189-193.
[5] LIEDTKE F F . Computer simulation of an automatic classification procedure for digitally modulated communication signals with unknown parameters[J]. Signal Processing, 1984, 6(4):311-323.
[6] SILLS, J.A. Maximum-likelihood modulation classification for PSK/QAM[C].Military Communications Conference Proceedings. IEEE . 1999.
[7] YAWPO YANG, JEN-NING CHANG, JI-CHYUN LIU, et al. Maximum Log-Likelihood Function-Based QAM Signal Classification over Fading Channels[J]. An Internaional Journal,2004,28(1):77-94.
[8] LOPATKA J , PEDZISZ M . Automatic modulation classification using statistical moments and a fuzzy classifier[C] International Conference on Signal Processing. IEEE, 2000.
[9] 贺涛,周正欧. 使用分集技术的信号调制类型识别[J]. 电子与信息学报,2008,30(4):872-875.
[10] SHAKRA, MAHMOUD M., et al. Automatic digital modulation recognition of satellite communication signals, National Radio Science Conference,2015 [C].Giza, Egypt.:Institute of Electrical and Electronics Engineers, 2015.
[11] K. C. HO, W. PROKOPIW, Y. T. CHAN. Modulation identification of digital signals by the wavelet transform[J]. IEE proceedings.Radar, sonar and navigation,2000,147(4):169-176.
[12] 刘建林. 基于高阶累积量通信调制信号类型的综合识别[J]. 舰船电子工程,2022,42(10):73-76.
[13] ZHANG, JUAN, LI, et al. Modulation classification method for frequency modulation signals based on the time-frequency distribution and CNN[J]. IET radar, sonar & navigation,2018,12(2):244-249.
[14] 李东瑾,杨瑞娟,董睿杰. 基于深度时频特征学习的雷达辐射源识别[J]. 国防科技大学学报,2020,42(6):112-119.
[15] 吴爱华,彭金喜. 基于深度残差收缩网络的信号调制类型识别[J]. 电子信息对抗技术,2022,37(4):24-30.
[16] O’SHEA T, CORGAN J, CLANCY T C.Convolutional Radio Modulation Recognition Networks[C].Cham,Switzerland:Springer,2016.
[17] 董章华,赵士杰,赖莉.一种基于Swin Transformer神经网络的低截获概率雷达信号调制类型的识别方法[J].四川大学学报(自然科学版),2023,60(02):42-48.
[18] O'SHEA T , ROY T, Clancy T C. Over the Air Deep Learning Based Radio Signal Classification[J]. IEEE Journal of Selected Topics in Signal Processing, 2017, PP(99):1-1。
[19] 李红领, 谢飞, 杜广超. 单音干扰下BPSK调制信号的最佳干扰性能分析.亚太地区信息论学术会议[C]. 西安:Scientific Research Publishing,2010.
[20] 杨豪, 颜青, 刘会来. 单音干扰对QPSK解调性能影响分析[J]. 无线电工程,2014,44(04):54-57.
[21] 王伟, 杨俊安, 刘辉. 基于干扰效果在线评估的行为参数分析[J]. 电子信息对抗技术,2016,31(06):76-80.
[22] 岑新龙, 张新如, 王少键. 基于灰色关联分析的通信干扰效果评估[J]. 舰船电子对抗,2011,34(3):83-87.
[23] 张会, 刘伯栋. 基于眼图分析的数字通信干扰效果方法研究[J]. 舰船电子工程,2011,31(02):73-76.
[24] 赵红梅, 惠守强, 王建等. 通信设备受扰性能评估方法研究[J]. 中国电子科学研究院学报,2013,8(04):398-402.
[25] 于少伟, 陈卫东, 张华冲. 单音干扰对16QAM解调性能影响分析[J]. 无线电工程,2015,45(01):30-32,64.
[26] CEM SEN. Digital Communications Jamming[D].Master of Science in Systems Engineering form the Naval Post-graduate School,2000.
[27] WICKERT, M.A.. Digital communication with jamming experiments for a signal processing first course. IEEE Digital Signal Processing Workshop [C].Sedona,USA:IEEE, 2011.
[28] 利强, 唐友喜. 单音干扰环境中M-QAM的性能分析[C].成都:四川省通信学会, 2007.
[29] 李志友, 李宏. 神经网络在通信干扰效果评估中的应用研究[J]. 舰船电子工程,2005(04):109-112.
[30] 裴毅, 刘春茂, 金卫同等. 时分噪声调频干扰性能分析[J]. 无线电工程,2010,40(12):12-20.
[31] 闫云斌, 全厚德, 崔佩璋. GMSK跳频通信跟踪干扰性能分析[J].电子技术应用,2012,38(05):109-112.
[32] 栗苹, 赵国庆, 杨小牛等. 信息对抗技术[M]. 北京: 清华大学出版社,2008.
[33] 刘力维, 魏成巍, 李勇等. 电子对抗训练中通信干扰效果评估方法研究[J]. 计算机与现代化,2010(11):77-79.
[34] LECUN Y, BOTTOU L. Gradient-based learning applied to document recognition[J].Proceedings of the IEEE, 1998, 86(11): 2278-2324
[35] GLOROT X, BENGIO Y. Understanding the difficulty of training deep feedforward neural networks[J]. Journal of Machine Learning Research, 2010, 9:249-256.
[36] He K , Zhang X , Ren S , et al. Delving Deep into Rectifiers: Surpassing Human-Level Performance on ImageNet Classification[J]. IEEE Computer Society, 2015,1026-1034
[37] 洪磊,杨育红,张瑞. QPSK最佳干扰研究与仿真[J]. 通信技术,2009,42(2):8-11.
[38] El-Mahdy E S . Multiple tone interference of multicarrier frequency-hopping BPSK system for a Rayleigh fading channel with channel estimation errors[J]. Digital Signal Processing, 2010, 20(3):869-880.
[39] 张容,何明浩,王振华. 通信干扰误码率分析与仿真[J]. 舰船电子对抗,2008,31(4):73-76.
[40] 邱丹,张彬,许华,等. 单频干扰对QPSK 载波同步影响的分析[J]. 信息安全与通信保密,2005(6):104-106.
[41] Vandendorpe L . Multitone system in an unlimited bandwidth multipath Rician fading environment[C]. Iee European Conference on Mobile & Personal Communications. IET, 1993.
[42] STEFANOVIC M, MILOSENIC MS,et al. Performance of qpsk system in the presence of pulse interference and noisy carrier reference signal[J]. Serbian Journal of Electrical Engineering, 2003, 1(1):123-130.
[43] 孙文友. QAM调制解调技术的研究和实现[D]. 黑龙江:哈尔滨工业大学,2005.
[44] 孙志国,徐天宇,邓昌青,等. 16-QAM信号的最佳干扰分析[J]. 哈尔滨工程大学学报,2018,39(7):1245-1250.
[45] SIMON M K . Differentially coherent detection of QASK for frequency-hopping systems. II - Performance in the presence of jamming[J]. IEEE Transactions on Communications, 1982, 30(1):165-172.