Research On Anti-jamming Method Of TT&C In Airspace Based On DOA Estimation Of Convolutional Neural Network

With the electromagnetic environment becoming more and more complex and various interferences becoming more and more serious,the anti-interference ability of the measurement and control system determines its viability in the complex electromagnetic environment.Spatial anti-jamming technology based on adaptive array antenna is always an important method of anti-jamming research in measurement and control system.After years of development,many achievements have been made in DOA estimation and beamforming of two parts of adaptive array antenna technology.However,DOA estimation still faces the problem of large amount of calculation and high precision,and the error of DOA estimation will directly affect the effect of beamforming.In this paper,the principle of array antenna and convolutional neural network are combined to study the fast and accurate DOA estimation method based on convolutional neural network,and then beam forming technology based on certain criteria is used to achieve the purpose of anti-jamming in airspace.Finally,the measurement and control system model is established and the anti-jamming efficiency is evaluated.Firstly,the basic principle of array antenna,the basic principle of direction estimation and beamforming,and the characteristics of convolutional neural network are described.Then a method of DOA estimation using convolutional neural network is proposed.The experimental results show that convolutional neural network can estimate DOA of single source and multi-source,which can achieve the same high-precision estimation as music algorithm,but the time is greatly shortened.Then a cnn-music joint estimation method is proposed,which can still be used when the SNR is as low as-20 d B In order to improve the speed of DOA estimation,convolutional neural network is used to reduce the peak searching range of music algorithm.Then,aiming at the problem of coherent sources,the smooth filtering method is used to decoherence,so that the direction of interference sources estimated by music algorithm is accurate.Then,the beamforming method is analyzed and compared.It is concluded that the beamforming method based on the linear constraint minimum variance criterion is most suitable for the measurement and control system in this paper.The reason is that the beamforming method based on the maximum SINR criterion and the beamforming method based on the linear constrained minimum variance criterion have a strong inhibition effect on the interference direction compared with the conventional beamforming method.The condition of the beamforming method based on the linear constrained minimum variance criterion is that we need to know the direction of the expected signal and the interference signal.This condition is that we can It can be estimated accurately by convolution neural network.Finally,the anti-jamming model of TT & C system in airspace is established.The DOA estimation method in Chapter 3 is used as a priori condition.Then,the beamforming method discussed in Chapter 4 is added to the model to discuss the anti-jamming ability of the anti-jamming system in airspace in the whole TT & C system.The dry signal ratio and error before and after using the antijamming system are established The change curve of the code rate and the anti-jamming performance of the simulation system are evaluated.