Equalization system of low differential mode delay few-mode fibers based on the neural network MIMO algorithm

Area of Science:

• Optical communications
• Fiber optics engineering
• Signal processing

Background:

• Information networks require higher transmission capacity, but single-mode fiber capacity is nearing its theoretical limit.
• Multidimensional multiplexing is a key technique to increase fiber capacity.
• High differential mode delay (DMD) in optical fibers complicates signal demultiplexing.

Purpose of the Study:

• To optimize a trench-assisted gradient refractive index fiber structure to reduce DMD.
• To develop an advanced equalization algorithm for high-capacity optical transmission systems.
• To verify the performance of the optimized fiber and algorithm in a simulated optical transmission system.

Main Methods:

• Intelligent design methods were used to optimize the trench-assisted gradient refractive index fiber structure.
• A novel least mean squares-feedforward neural network constant modulus algorithm (LMS-FNNCMA) was designed.
• A polarization division multiplexing-wavelength division multiplexing-mode division multiplexing (PDM-WDM-MDM) optical transmission system was simulated.

Main Results:

• The optimized optical fiber structure achieved a maximum DMD of 19.6 ps/km.
• The LMS-FNNCMA algorithm successfully mitigated crosstalk in the simulated system.
• The system demonstrated successful de-crosstalk over a 1200 km transmission distance at 1.2 Tbps.

Conclusions:

• The optimized trench-assisted gradient refractive index fiber structure effectively reduces DMD.
• The developed LMS-FNNCMA algorithm is effective for equalization in complex optical transmission systems.
• The study demonstrates a viable approach for achieving ultra-high-speed optical communication.