Research On Intelligent Communication Jamming Decision-making Technology Based On Reinforcement Learning

With the development of artificial intelligence,more and more attention has been paid to the cognitive electronic warfare technology which combines artificial intelligence with electronic warfare.Intelligent communication jamming decision-making is an important research direction of cognitive electronic warfare.Intelligent communication interference decision-making is essentially a decision-making problem,and reinforcement learning is a feasible algorithm for sequential decision-making.Thus,this paper studies the intelligent communication interference decision-making algorithm based on reinforcement learning.Firstly,the basic principle and algorithm framework of reinforcement learning are studied.Several classical learning algorithms of reinforcement learning are given.Then,the deep feedforward network is introduced,and its training method is analyzed.Combined with deep learning and reinforcement learning,the concept of deep reinforcement learning and several typical algorithms are introduced.Then this paper introduces the cognitive communication countermeasure system,and analyses the position of intelligent communication jamming algorithm in the system.Then three application scenarios of intelligent jamming algorithms,such as invariable communication parameters of the enemy,channel switching of the enemy according to the interference situation,and communication parameters changing according to the interference situation of the enemy,are proposed.For each application scenario,the appropriate intelligent interference decision algorithm is proposed.The reward function,algorithm flow and algorithm parameters of the algorithm in each application scenario are designed.Aiming at specific problems,the algorithm is improved and its learning process is optimized.Finally,according to the above analysis results,the simulation environment of intelligent jamming is built.Intelligent jamming algorithms in three application scenarios are validated.The simulation results show that the algorithms can learn good jamming strategies in corresponding application scenario.The result proves the feasibility of intelligent jamming algorithm in communication countermeasure.