We design an optical feedback network making use of machine learning techniques and demonstrate via simulations its ability to correct for the effects of turbulent propagation on optical modes. This artificial neural network scheme only relies on measuring the intensity profile of the distorted modes, making the approach simple and robust. The network results in the generation of various mode profiles at the transmitter that, after propagation through turbulence, closely resemble the desired target mode. The corrected optical mode profiles at the receiver are found to be nearly identical to the desired profiles, with near-zero mean square error indices. We are hopeful that the present results combining the fields of machine learning and optical communications will greatly enhance the robustness of free-space optical links.
[1]
Md. Rafiqul Islam,et al.
Investigating of rain attenuation impact on Free Space Optics propagation in tropical region
,
2011,
2011 4th International Conference on Mechatronics (ICOM).
[2]
R. Scarmozzino,et al.
Simulation-assisted design of free space optical transmission systems
,
2005,
MILCOM 2005 - 2005 IEEE Military Communications Conference.
[3]
Zach DeVito,et al.
Opt
,
2017
.
[4]
Szymon Gladysz,et al.
Adaptable adaptive optics and image processing at Fraunhofer IOSB
,
2015
.
[5]
Charu C. Aggarwal,et al.
Neural Networks and Deep Learning
,
2018,
Springer International Publishing.