Ultra-fast and accurate multimode waveguide design based on a dataset-based eigenmode expansion method.

Area of Science:

• Photonics and Optical Engineering
• Computational Electromagnetics
• Materials Science (Silicon Photonics)

Background:

• Designing multimode waveguides is computationally intensive using traditional methods.
• Accurate and efficient simulation tools are crucial for advancing photonic integrated circuits.
• Existing simulation techniques often lack the speed required for complex device optimization.

Purpose of the Study:

• To introduce a novel dataset-based photonic simulation framework for multimode waveguide design.
• To achieve ultra-fast and highly accurate simulations for complex waveguide structures.
• To enable rapid design and optimization of large-scale multimode photonic devices.

Main Methods:

• Development of a dataset-based simulation framework.
• Implementation of the framework for silicon multimode waveguide bend design.
• Validation of simulation accuracy against a commercial 3D finite-difference time-domain (FDTD) method.

Main Results:

• Achieved a speed-up of two to three orders of magnitude compared to conventional methods.
• Successfully designed a silicon multimode waveguide bend (30 µm effective radius) in under one second.
• Demonstrated efficient device optimization, designing bends with arbitrary power splitting ratios in 68 minutes on a CPU.

Conclusions:

• The proposed framework significantly accelerates multimode waveguide design and simulation.
• It offers high accuracy, comparable to established commercial methods.
• This approach democratizes computationally intensive photonic simulations for designers.