Modulation Classification Method based on Deep Learning under Non-Gaussian Noise

The arrival of 5G has accelerated the development of the Internet of things and vehicular technology, which often need to transmit large amounts of data through wireless networks. Modulation classification plays an important role in wireless communication. Recent years, deep learning has been applied to solve the modulation classification problem and achieved good classification results. At present, almost all the papers that use deep learning to solve modulation classification are in Gaussian White noise environment. However, the error source mainly comes from non-Gaussian noise in practical wireless communication. In this paper, a modulation classification method in non-Gaussian environment based on Deep Learning is proposed. The proposed algorithm can effectively suppress the sharp pulse in non-Gaussian noise and improve the modulation recognition accuracy. MPSK and MQAM signals which are difficult to distinguish are adopted in the simulation experiment. The simulation results show that validity of the proposed method. At the same time, experiments show that this method is robust to the characteristic exponent of noise.