Adaptive Demodulation Technique for Efficiently Detecting Orbital Angular Momentum (OAM) Modes Based on the Improved Convolutional Neural Network

Abstract

Convolutional neural network (CNN), as a model of deep learning (DL), has been widely applied to the field of computer vision as well as optical communication. In this paper, we focus on the adaptive demodulation technique in orbital angular momentum (OAM) free-space optical (FSO) communication system with the improved CNN. In order to achieve adaptive demodulation under free-space turbulence channel, the traditional CNN with our preliminary optimization methods has been firstly demonstrated. Then, in view of the relatively poor performance of traditional CNN for OAM modes demodulation under strong turbulence environment with long transmission distance, the network was deepened and the residual learning framework was used to solve the degradation problem. We have respectively investigated the performance of this improved CNN by testing the demodulation performance in 4-ary, 8-ary, 10-ary and 16-ary OAM systems, and analyzed the generalization ability of accommodating unknown turbulence environment by training the model with different training sets. The numerical simulation shows that the demodulation accuracy is about 100.0%, 99.5%, 99.2% and 99.0% respectively for 4-ary, 8-ary, 10-ary and 16-ary OAM systems over the 2000m free-space with strong turbulent level. And our improved CNN trained by a hybrid training set with several levels of turbulence can provide the model with stronger ability to accommodate more kinds of unknown turbulence environments. It is anticipated that these results might be helpful for improving the reliability of the OAM-FSO communication system in future.