Abstract
In an orbital angular momentum-shift keying free-space optical (OAM-SK FSO) communication system, precisely recognizing OAM superposed modes at the receiver site is crucial to improve the communication capacity. While deep learning (DL) provides an effective method for OAM demodulation, with the increase of OAM modes, the dimension explosion of OAM superstates results in unacceptable costs on training the DL model. Here, we demonstrate a few-shot-learning-based demodulator to achieve a 65,536-ary OAM-SK FSO communication system. By learning from only 256 classes of samples, the remaining 65,280 unseen classes can be predicted with an accuracy of more than 94%, which saves a large number of resources on data preparation and model training. Based on this demodulator, we first realize the single transmission of a color pixel and the single transmission of two gray scale pixels on the application of colorful-image-transmission in free space with an average error rate less than 0.023%. This work may provide a new, to the best of our knowledge, approach for big data capacity in optical communication systems.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Disclaimer: 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.