Research On Methods Of Radar Mainlobe/Near Mainlobe Jamming Suppression

In modern warfare,radar has the ability to obtain enemy information and perceive battlefield environment,so it plays an important role.The birth of electronic countermeasure technique has seriously threatened the working performance and survivability of radar.With the development of electronic warfare techniques,the problem of complex electromagnetic environment is becoming increasingly prominent,especially the mainlobe / near mainlobe jamming entering from the main beam of radar,which brings severe challenges to the detection and tracking of radar system,and puts forward higher requirements for the antijamming method of radar.Aiming at the problem of mainlobe / near mainlobe jamming suppression and improving the anti-jamming capability of radar,this paper studies the suppression methods of mainlobe / near mainlobe jamming combined with frequency agility and adaptive sum-difference beamforming.The main work of this paper is summarized as follows:1.Aiming at the problem of mainlobe / near mainlobe jamming suppression,the spatial near mainlobe jamming suppression methods and mainlobe jamming suppression methods in frequency domain are studied.Firstly,aiming at the near mainlobe jamming scene,three spatial near mainlobe jamming suppression algorithms are introduced,including blocking matrix preprocessing,eigenprojection matrix preprocessing and adaptive sum-difference beamforming.The performance of jamming suppression of the three methods is simulated and analyzed,and the advantages,disadvantages and limitations of the three methods are pointed out respectively.Secondly,aiming at the mainlobe jamming scene,the principle of frequency agility against jamming and three forms of frequency agility are introduced,and the advantages and disadvantages of frequency agility radar are analyzed.The simulation shows that frequency agility radar has the ability to resist mainlobe jamming compared with pulse doppler radar with fixed carrier frequency.2.Aiming at the problem of near mainlobe jamming suppression in multi-jamming scene,jamming suppression methods by using sum and difference beams in 4-channel based on area array model under multiple near mainlobe interferences scene and mainlobe and sidelobe interferences scene are studied.Firstly,aiming at the scene of multiple near mainlobe interferences,this paper studies an anti-jamming method based on expanding the degrees of freedom of difference beams.This method increases the numbers of multiple difference beams,so as to provide enough degrees of freedom to suppress multiple interferences.The experimental results show that this method can suppress multiple near mainlobe interferences at the same time.Secondly,aiming at the scene where mainlobe jamming and sidelobe jamming exist at the same time,this paper studies two partially adaptive anti-jamming methods at subarray level and their corresponding angle measurement methods: jamming suppression methods based on sidelobe cancellation algorithm,generalized sidelobe cancellation algorithm with sum and difference beams.Both methods add an auxiliary channel pointing to the sidelobe jamming on the basis of 4 channels.While suppressing the near mainlobe jamming in fractal dimension,the auxiliary channel is used to suppress the sidelobe jamming in sum beam.The simulation results show that both methods can suppress the mixed jamming effectively and measure the angles of target accurately,and the latter performs better on angle measurement accuracy.3.Aiming at the problem of mainlobe jamming suppression under the frequency agile radar model,a learning method of anti-jamming strategy of frequency agile based on deep reinforcement learning is proposed.Frequency agile radar has a variety of frequency hopping modes,and different frequency hopping strategies will affect the anti-jamming performance and detection performance of the radar.Therefore,the learning of frequency hopping strategy is of great significance to improve the ability of mainlobe jamming suppression of frequency agile radar.This paper establishes the Markov decision process(MDP)problem model of the competition between radar and jammer,and designs the key factors such as actions,states and rewards.Proximal policy optimization algorithm is used to solve the MDP problem mentioned above under the actor-critic framework with attention mechanism.It can be seen from the simulation results that this method is capable of learning the efficient antijamming strategy against a specific jamming strategy and improving the anti-jamming performance.In addition,aiming at the situation that the jammer adopts a variety of different jamming strategies,a unified anti-jamming strategy learning method based on policy distillation algorithm is studied.According to this method,multiple anti-jamming strategies are migrated to a deep neural network to obtain the unified anti-jamming strategy.The experimental result shows that this method enables the radar to combat multiple jamming strategies at the same time without losing the performance of the original anti-jamming strategy.