Research On Modulation Pattern Recognition Of Radio Signals Based On Machine Learning

With the rapid development of artificial intelligence technology,deep learning algorithm has become the main research direction in the field of machine learning.The deep learning model,represented by convolutional neural network and recurrent neural network,creatively excavates the deep characteristics and timing characteristics of radio signals at the level of data science,and realizes the modulation pattern recognition of radio signals on this basis.In this paper,three improved convolutional neural networks(myNet,myNet2,myVGGNet)are proposed.Compared with the classical convolution neural network(CNN),the improved convolution neural network(myNet)can achieve the performance gain in the modulation pattern recognition process by reducing the number of convolution kernels and increasing the number of convolution layers.The improved convolution neural network(myNet2)is optimized in the structure model,and compared with the classical convolution neural network,by adding a long-term and short-term memory networks(LSTM)into the network structure,the convolution layer is used to extract the deep features of radio signals,and the long-term and short-term memory network layer is used to extract the temporal correlation of radio signals,so as to achieve the performance gain in the modulation pattern recognition process.The improved convolution neural network(myVGGNet)is optimized in the structure model and pooling method,and compared with the classical convolution neural network,the structure model adopts the structure of sequential convolution module and full connection layer,the pooling method changes the pooling method in the fourth sequential convolution module of the improved convolution neural network,the average pooling method is used to improve the overall fitting ability of the network to achieve the performance gain in the modulation pattern recognition process.The simulation results show that the improved convolutional neural network(myNet,myNet2,myVGGNet)can generate some performance gains in the modulation pattern recognition of radio signals.So as to promote the application of machine learning algorithm in the field of modulation pattern recognition of radio signals.