All-optical image transmission through a dynamically perturbed multimode fiber and a ring-core fiber using...

Area of Science:

• Optics and Photonics
• Artificial Intelligence
• Telecommunications

Background:

• Fiber optic imaging faces challenges with signal scattering, particularly in multimode fibers (MMF) and ring-core fibers (RCF).
• Traditional methods for correcting scattering-induced distortions often have limitations in complex field reconstruction.
• Orbital angular momentum (OAM) modes offer potential for enhanced information capacity in optical communication and imaging.

Purpose of the Study:

• To introduce and experimentally validate an all-optical image reconstruction technique using a diffractive deep neural network (D2NN).
• To leverage Orbital Angular Momentum (OAM) modes for efficient imaging transmission through optical fibers.
• To address and overcome the scattering-to-restoration transformation issues inherent in RCF.

Main Methods:

• Implementation of a diffractive deep neural network (D2NN) integrated within a ring-core fiber (RCF) architecture.
• Utilization of Orbital Angular Momentum (OAM) modes for propagating image data.
• Experimental validation at a 1550 nm operating wavelength, comparing performance in both MMF and RCF.

Main Results:

• Successful experimental demonstration of complex field diffractive image reconstruction.
• Significant mitigation of scattering-induced artifacts within RCF.
• Demonstrated superior performance of the D2NN-RCF approach compared to conventional MMF-based correction techniques.

Conclusions:

• The proposed D2NN-based all-optical image reconstruction technique is effective for RCF imaging.
• This method offers a significant advancement in overcoming fiber scattering challenges.
• The approach shows promise for high-fidelity imaging transmission in complex fiber environments.