Research On Phased Array Radar And Decoy Recognition Method Based On Spatial Polarization Characteristics

For fully polarimetric phased array radar,its array element has two ports that can feed independently.By changing the amplitude and phase of port excitation,controlling the polarization state of beam and controlling the energy distribution of radiation field,it has a wide range of military applications.The active decoy system is deployed around the fully polarimetric phased array radar to effectively protect the phased array radar.In order to effectively attack phased array radar,it is necessary to identify fully polarimetric phased array radar and active decoy.Due to the polarization characteristics as the inherent attribute of the antenna and the differences in the spatial polarization characteristics between the fully polarimetric phased array radar and the active decoy,this paper studies the spatial polarization characteristics of the two,and proposes a new identification method between the fully polarimetric phased array radar and the active decoy combined with deep learning.The main content is divided into three parts:In the first part,the spatial polarization characteristics of antenna are studied.Because the spatial polarization characteristics of phased array radar and decoy antenna are related to the spatial polarization characteristics of its array elements,it is necessary to study the spatial polarization characteristics of antenna elements and the factors affecting the polarization characteristics of antenna elements.Firstly,the expressions of different polarization types of electromagnetic wave are deduced,the commonly used antenna polarization characteristic parameters are introduced,and the corresponding relationship between polarization characteristic parameters and polarization types is given.Then,the polarization characteristics of some common array elements are analyzed theoretically and simulated by HFSS electromagnetic simulation software,and the factors affecting the spatial polarization characteristics of the antenna are studied.Finally,through theoretical analysis and electromagnetic calculation,the spatial polarization characteristics of the ideal fully polarimetric phased array antenna with microstrip antenna as the array element and the spatial polarization characteristics of the pyramid horn antenna as the active bait are obtained.The spatial polarization characteristics of the ideal fully polarimetric phased array antenna are compared with those of the array element,it provides a basis for follow-up research.The second part is the modeling and analysis of spatial polarization of full polarization phased array radar.At present,phased array radar only considers beamforming modeling or polarization modeling without considering mutual coupling effect,and there are few modeling methods considering the combination of the two.Therefore,this paper constructs a fully polarimetric phased array radar model combining mutual coupling compensation and polarization control.Firstly,the mutual coupling analysis and correction method of fully polarimetric phased array radar are studied.The mutual coupling coefficient is obtained by using the least square method,and the excitation of different array elements is adjusted to make the spatial polarization characteristics of fully polarimetric phased array radar in the ideal spatial polarization characteristics,which is convenient for polarization control.Then,the beamforming method of fully polarimetric phased array radar is studied.The phase of excitation is adjusted to realize beam scanning.At the same time,Taylor weighting method is used to change the excitation proportion and reduce the side lobe level to meet the requirements of low side lobe.Finally,the polarization control method of full polarization phased array radar is studied,the spatial polarization modeling of full polarization phased array radar is realized,and the spatial polarization characteristics are obtained.Through experiments,the feasibility of obtaining the mutual coupling coefficient by the least square method and constructing the full polarization phased array radar model combining mutual coupling compensation and polarization control are verified.The experimental results show that the polarization characteristic parameters of full polarization phased array radar and active decoy are different in the same airspace,which provides a theoretical basis for the later identification part.In the third part,a new identification method of full polarization phased array radar and active decoy is studied.Firstly,the training data set and test data set required by the model are established,and the size and construction mode of the data set are described in detail.At the same time,the preprocessing steps are described in detail.Combining convolution and Long Short Term Memory neural network(LSTM),this paper proposes LSTM-FCN model.The feature is extracted by LSTM module and convolution module to solve the problem of recognition between full polarization phased array radar and active decoy.In this paper,the model is compared with convolution network model and long-term and short-term memory neural network model.At the same time,the effects of different SNR training data sets,data adoption methods and model iteration times on the recognition accuracy are studied.The experimental results show that the LSTM-FCN network model proposed in this paper can effectively identify the fully polarimetric phased array radar and active decoy.When the signalto-noise ratio is-1d B,the recognition accuracy is 86.45%.