A Deep Learning based Detector for FSO System Considering Imperfect CSI Scenario

In free space optical (FSO) communication system, the maximum likelihood (ML) is the optimum detector considering perfect channel state information (CSI) at receiver. However, the ML does not provide a proper performance in imperfect CSI scenario. Therefore, a deep learning based detection method is proposed in this paper. In this paper, we use deep neural network to get a mapping function of a received signal and transmitted symbol streams. Moreover, the end-to-end approach using deep learning with the conventional ML method is compared for both perfect and imperfect scenarios. Simulation results show that our method presents a better performance compared to the conventional ML method in imperfect CSI. Moreover, our proposed method reaches the same performance as the ML detector in perfect CSI.