Research On Intelligent Anti-jamming Decision Based On Hidden Idea

In recent years,with the rapid development of wireless communication technology,the application range involves all walks of life,and the number of communication users is increasing rapidly.Wireless communication has brought a lot of convenience to people's life,but also produced some urgent problems.Hostile jamming will attack the user's communication,and the unintentional jamming between users will also affect the user's experience.It can be imagined that if there is no corresponding anti-jamming technology to solve these problems,the quality and safety of wireless communication will not be guaranteed.Therefore,it is of great significance to study the anti-jamming technology of communication.The new anti-jamming technology fully considers the dynamic characteristics of the environment,the decision changes with the change of the environment,and the ability to deal with intelligent jamming is significantly improved.However,this kind of method generally takes the idea of avoidance as the core,that is,to optimize the instantaneous or long-term communication effect as the goal,and does not directly care about the user's signal waveform or decision rules being captured by the jammer.However,in the real anti-jamming problem,communication and jamming are often not one-time confrontation.The information leakage of user in one confrontation may lead to the next more targeted jamming waveform or jamming decision.Therefore,in the long run,anti-jamming technology should not only ensure the communication effect,but also reduce the information leakage as much as possible.In this paper,we consider that there are some differences between communication links and jamming links due to channel fading or environmental signals.If the communication signal can be hidden to avoid jamming and obtain user information,the jammer can't carry out targeted attacks.In this paper,based on the idea of concealment and deep reinforcement learning algorithm,different jammer modes are studied.First of all,when the jamming decision can be measured,this paper designs a method to measure the correlation of decision sequence,and judges whether the jammer has obtained the user's decision through the correlation of decision.Combined with deep reinforcement learning algorithm,a hidden anti-jamming method based on decision correlation is proposed.The simulation results show that it can't only improve the communication efficiency,but also avoid the jammer of user information acquisition.Unfortunately,when the jammer decision cannot be measured or the jammer decision and the user decision are not in the same dimension,the above method will not be applicable.Jammer's decision is an attack based on user's decision,and the change of jammer decision will cause the corresponding change of environment state.In this paper,a user decision evaluation network is designed based on generative adversarial network.According to the change of environment state,whether the user decision is interfered or not is deduced.Finally,combined with deep reinforcement learning algorithm,a hidden anti-jamming method based on environment reconstruction is proposed.Simulation results show that it can deal with more complex jammer mode and ensure the security of user decision information.