Research On Intelligent Detection Method Of Emitter Abnormal Target In Complex Electromagnetic Environment

In modern warfare,electronic reconnaissance is a common equipment used to obtain military intelligence,and the abnormal detection of emitter is its key technology.It has great prospects in combating enemy interference and unknown emitters.With the rapid development of electronic technology and electronic countermeasures,the electromagnetic environment has gradually become complicated.Current traditional methods are limited by the low signal noise ratio(SNR)and complex noise suppression process,and managing dynamic electromagnetic environments is difficult.Therefore,a breakthrough in traditional processing methods is warranted to further improve the detection performance of the abnormal emitter.Under such conditions,how to effectively realize the intelligent detection of electronic reconnaissance has valuable research significance.The paper studies the problem of the intelligent detection of abnormal emitter in a complex environment.The specific research contents are as follows:(1)The basic characteristics of the complex electromagnetic environment and the basic theory of intelligent detection are introduced.First,a complex electromagnetic environment model is established on the basis of multi-source heterogeneous signal composition,high pulse density,and multi-path reflection,and the corresponding mathematical expressions are summarized.Second,the limitations of traditional detection methods are analyzed in complex electromagnetic environments.Finally,this paper studies the related basis theories of intelligent detection on the basis of deep learning,and focuses on the structural framework of the relevant network model,which provides theoretical support for the subsequent research work.(2)Given that the signal detection is limited by low SNR,on the basis of the idea of signal and noise integration,an electromagnetic signal intelligent detection method on the basis of integration features is proposed.On the basis of the signal's intra-pulse characteristic information,the signal detection problem is transformed into a similar image classification problem with the integration feature of the electromagnetic signal,and then combined with convolutional neural network(CNN)to achieve abnormal detection.The simulation results show that compared with the traditional energy detection method,this method has better detection performance in developing the traditional approach of suppressing noise which is limited by low SNR.(3)To further enhance the intra-pulse characteristics of electromagnetic signals,on the basis of the idea of time reversal,an electromagnetic signal intelligent detection method on the basis of focused integration features is proposed.The time reversal technology is introduced to develop the integrated energy characteristics of the signal and combined with the image classification algorithm to complete the intelligent detection of abnormal emitter target in complex electromagnetic environments.The simulation results unveil that this method can effectively enhance the signal's intra-pulse characterization while improving the detection performance and reliability.(4)To solve the anomaly detection problem of electromagnetic batch targets for which parameter estimation has been completed,an electromagnetic signal intelligent detection method on the basis of inter-pulse information is proposed.The long short-term memory(LSTM)network prediction model is applied to the abnormal emitter detection on the basis of real-time pulse description word(PDW).On the basis of the emitter signal simulator,the method is verified and the anomaly confidence is introduced to combine unsupervised learning with manual review to realize real-time anomaly detection in the complex electromagnetic environment.The simulation results reveal that the method can effectively identify the abnormal target of emitter and further improve the robustness of the prediction model.