Abstract
In this paper, the deep learning-based Belief Propagation (BP) decoding is first investigated in a polar-coded free space optical (FSO) communication system. On the basis of neural Cyclic redundancy check-Polar Belief Propagation (NCPBP) decoding scheme, we propose a dual-recurrent neural network (RNN) based decoder for polar codes, which adopts a new training loss function, RNN structure and tanh-modified input. The presented decoder exhibits better performance to the NCPBP decoder under different intensities of turbulence. Furthermore, the decoder trained under fixed turbulence intensity shows higher turbulence adaptability and its advantage increases with the intensity of turbulence, reaching a decoding gain of 1 dB under strong turbulence when Rytov variance is 3.5.
Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
