Research On Intelligent Jamming Decision And Method Of UAV Group Cooperation Against Ground-to-air Radar

In the modern battlefield,ground-based anti-air radar is responsible for reconnaissance,tracking,fire targeting and other tasks,which seriously threatens our air power and is the key target of electronic countermeasures.As one of the most popular battlefield weapons in recent years,UAV can undertake a series of tasks,such as strike,reconnaissance and jamming.Especially in Ukraine in 2022,UAV plays a huge role.UAV has the characteristics of low cost,small size,flexible control and can avoid casualties.It is an excellent jammer carrier.There are two problems in the cooperative jamming of UAVs to radar,that is,how to select the appropriate jamming method and how to realize the jamming method.In order to solve these problems,this thesis is divided into three parts to discuss how to carry out cooperative jamming,how to select jamming methods and how to optimize jamming methods.This thesis analyzes the scenario of multiple UAVs fighting single radar formed after resource allocation,studies how to realize accurate and detailed jamming,and discusses the cooperative jamming method and intelligent jamming decision of UAV group.The main contents of this thesis are as follows:1.In this thesis,the cooperative jamming method of UAVs is analyzed.According to different effects,it can be divided into two types: improving the traditional jamming methods and implementing new jamming methods.Aiming at cooperative noise suppression jamming,cooperative false target jamming,cooperative blinking jamming and cooperative phantom track deception jamming,the effectiveness of the jamming method under the condition of multiple UAVs is verified by simulation based on the analysis of basic principles and mathematical models.2.In this thesis,the intelligent selection of jamming method is studied on the premise of the research of jamming method and the reasonable modeling of battlefield environment.Considering the non-cooperative characteristics of radar in practice,DQN algorithm is used to realize the intelligent selection of jamming methods.By using the self-learning ability of reinforcement learning and the fitting ability of neural network,DQN can solve the problem of how to select jamming method intelligently.Even if the parameters describing the state have errors,the agent after learning can still choose a better jamming method.3.On the basis of studying the intelligent selection of jamming method based on DQN,this thesis studies the shortcomings of DQN algorithm,some improved algorithms are proposed,such as improved DQN algorithm based on dynamic expert database,D3 QN algorithm,transfer reinforcement learning algorithm and layered reinforcement learning algorithm.The improvement of DQN algorithm improves the convergence performance of the algorithm,and the conclusion is verified by simulation4.This thesis optimizes intra-group collaboration of UAVs.In view of the broadband noise suppression jamming,the improved genetic algorithm and the improved cuckoo algorithm are respectively used to optimize the UAV’s task allocation strategy to achieve detailed jamming and maximize the jamming effect with limited jamming resources.