Research on Adaptive Communication Based on Recognition on Improving Convolutional Neural Network
Although wireless communication is widely used,the wireless communication environment is becoming more and more complex,and it is easily affected by various electromagnetic interference and e-lectromagnetic attack.Improving its anti-interference performance has always been the focus of wireless communication.To solve this problem,a spectrum adaptive communication transmission method based on improving convolutional neural network is proposed by using the results of electromagnetic spectrum sens-ing of cognitive radio.Firstly,the traditional convolutional neural network and the improved convolutional neural network are used to identify the typical interference types,and the corresponding interference pa-rameters are estimated according to the identification results.On this basis,combined with the spectrum sensing results,the spectrum that is not affected by interference is adaptively selected for communication transmission according to the principle of efficient spectrum utilization efficiency,so as to achieve the du-al purpose of anti-interference communication transmission and improving spectrum utilization efficiency.The simulation results show that the recognition accuracy of the proposed method for interference types is more than 98％,and the mean absolute percentage error of interference parameter estimation is less than 0.01.Finally,the actual spectrum adaptive communication transmission experiment is carried out on the software radio platform,and the normal communication can be achieved under different signal to interfer-ence ratios.The experimental results show that the anti-interference communication effect is remarkable.
NETL
NSTL
万方数据
