Research On Intelligent Anti-jamming Technology Of Frequency Hopping Communication

Frequency hopping communication is widely used in military and civil applications because of its excellent anti-jamming,anti-interception and networking capabilities.However,with the continuous intelligence of electromagnetic jamming,the anti-jamming performance of traditional frequency hopping communication system cannot meet the actual needs.On the basis of realizing the perception of electromagnetic jamming environment,intelligent anti-jamming communication ensures the efficient and reliable information transmission of the system.At present,it has become a new research highlight in the field of anti-interference communication.This thesis mainly studies the intelligent anti-jamming technology of frequency hopping communication.Firstly,in order to improve the anti-jamming ability of frequency hopping communication system in electromagnetic jamming environment,an intelligent decision engine of frequency hopping communication system based on improved genetic algorithm is proposed.Aiming at the slow convergence speed and "premature" phenomenon of genetic algorithm,the chromosome similarity function is introduced in the selection operation to avoid the "premature" phenomenon and ensure the diversity of offspring population.The competitive strategy is adopted in the crossover operation,that is the crossover operation is carried out on the better individuals to speed up the convergence speed of the algorithm.In the mutation operation,the effective information of some excellent individuals is used to guide the mutation operation,so that the individual mutation has directionality,and the individuals that may be locally optimal are chaotically disturbed to reduce the probability of converging to the local optimal.Compared with adaptive genetic algorithm,the improved genetic algorithm significantly improves the convergence speed and the probability of converging to the optimal solution.The improved genetic algorithm is used as the decision algorithm of the intelligent decision engine to intelligently select the communication parameters such as modulation mode,transmission power and frequency hopping pattern in different jamming environments.The simulation results show that the intelligent decision engine using the improved genetic algorithm can effectively improve the anti-jamming ability of frequency hopping communication system.Secondly,by using the frequency hopping characteristics of frequency hopping communication and an anti-jamming frequency hopping sequence design method considering Hamming correlation,uniformity,jamming probability and frequency hopping gain of frequency hopping sequence is proposed with the combined jamming information.The frequency hopping sequence design problem is modeled as a single objective optimization problem,which is solved by grey wolf optimizer.A new reverse learning strategy is used in the grey wolf optimizer,which has a higher probability to find the optimal solution than the basic reverse learning,and accelerate the convergence speed of the optimizer.The information exchange mechanism using Levy flight is introduced to improve the information utilization rate of wolves and effectively avoid the phenomenon of convergence to the local optimal solution caused by the grey wolf commanding the wolves.Simulation results show that the improved grey wolf optimizer has faster convergence speed and better solution accuracy than other grey wolf optimizers in terms of basic test functions.In a variety of jamming situations,the frequency hopping sequence designed by this method has stronger anti-jamming ability than the traditional frequency hopping sequence and the wide interval frequency hopping sequence.Finally,in order to deal with the situation that the frequency hopping pattern is no longer fixed but intelligently selected in the intelligent frequency hopping communication system,an intelligent receiving scheme of frequency hopping signal is proposed.According to the different characteristics of frequency hopping signal and jamming signal in time-frequency domain,a CNN-GRU network is designed,which combines time-frequency analysis and deep learning to realize the intelligent estimation of frequency hopping sequence.The squeeze-and-excitation(SE)module is used in CNN-GRU to improve the feature extraction and expression ability of convolutional neural networks(CNN)for the time-frequency matrix of received signals.Gate recurrent unit(GRU)is used to improve the adaptability of the network to frequency hopping signals with different lengths.Finally,a classifier with a small number of categories is used to solve the problem that the network parameters are too large to converge.The simulation results show that the designed CNN-GRU network has strong generalization ability and robustness.On the premise of small samples,transfer learning is used to improve the adaptability and efficiency of the network model,and the bit error rate performance of the intelligent reception scheme is close to the reception performance under ideal conditions.