Convolutional Neural Network for Performance Monitoring and Modulation Formation Recognition in Communication System

In this paper, a deep learning-based method is proposed for system damage identification and modulation format recognition in satellite communication system. Four typical damages are studied, including atmospheric channel impairments and hardware imperfections. In addition, total of ten modulation formats are generated, from low-order phase modulation to complex quadrature amplitude modulation. All the data are converted into image data and processed from the perspective of image recognition using convolutional neural network without feature engineering.