With the communication environment becoming more and more complex,the demand for countermeasures is constantly increasing in civil and military communications.In order to achieve more intelligent and real-time anti-jamming,the intelligent decision-making technology of cognitive anti-jamming communication systems is studied in this thesis.In this thesis,cognitive anti-jamming intelligent decision-making engine is put forward.The engine includes three units with different functions: search decision-making unit,fast decision-making unit and modification decision-making unit.This thesis focuses on the ABC(Artificial Bee Colony Algorithm)based cognitive anti-jamming fast decision-making engine and the BP(Back Propagation)NN(Neural Network)based cognitive anti-jamming fast decision-making engine.The anti-jamming performance is simulated and analyzed in several typical environments.The details are as follows:In the first part,the background and significance of this thesis are introduced firstly,and the research status of cognitive anti-jamming technology and related technologies of intelligent decision-making are studied.The second part proposes intelligent decision-making engine of cognitive anti-jamming communication system.Based on the introduction of the existing CC(Cognitive Cycle)and cognitive engine model,the intelligent decision-making engine of cognitive anti-jamming communication system is proposed firstly.The model is designed and the knowledge expression is defined.Then a simple design of the search decision-making unit is done because it is not the focus of this article.In the end of this section studied the related algorithms of the fast decision-making unit and the modification decision-making unit.After introducing the principle of each algorithm,they are compared and selected theoretically.The third part studies the ABC based cognitive anti-jamming decision-making engine.Firstly,a decision framework and model are designed.Then the parameters such as initialization method,number of initial population,neighborhood search method,number of investigation bee and elimination threshold are selected by simulation.Finally,the anti-jamming performance of the ABC based cognitive anti-jammingdecision-making engine is simulated.And the simulation results show that the engine has strong anti-jamming alility in different interference types,different interference power,and different channel environments.The fourth part studies the BP-NN based cognitive anti-jamming fast decision-making engine.Firstly,the principles of the BP-NN—gradient descent method and back propagation method—are introduced.And the model of the engine is designed.Then the effects of network size,initialization method,number of small-batch data,and learning rate on network performance are verified by simulation and the most suitable parameters are selected.Finally,the good network performance and great anti-jamming capability of the BP-NN designed in this thesis is verified through simulation.At the same time,the advantages and disadvantages of the ABC and the BP-NN are compared.The ABC requires a search process for each decision-making,which takes more time.The BP-NN can realize real-time communication on the line and has a degree of fault tolerance and generalization ability,but the training sample error cannot be too large.The fifth part studies the overall performance of intelligent decision-making engine of cognitive anti-jamming communication system.Firstly,the rule based cognitive anti-jamming modification decision-making unit is introduced,and the intelligent decision-making algorithms used in this unit are studied.Then the three modules of the decision-making engine are combined to form a closed-loop intelligent decision-making communication system.Simulations are performed for different scenarios that may be encountered to illustrate the great anti-jamming capability of the system. |