کنفرانس مهندسی برق ایران

صفحه اصلی / سی و چهارمین کنفرانس بین المللی مهندسی برق

Python-Based Optimization of Inverse Taper Structures for Enhanced Fiber-to-Chip Coupling Efficiency Using FDTD Simulations

نویسندگان :

Soroush Rostami¹ Sarmad Rostami² Ali Poureslami³ Davood Ranjbar Rafi⁴

1- پژوهشگاه ارتباطات و فناوری اطلاعات 2- دانشگاه شهید بهشتی 3- پژوهشگاه ارتباطات و فناوری اطلاعات 4- پژوهشگاه ارتباطات و فناوری اطلاعات

کلمات کلیدی :

Photonic integrated circuits،fiber-to-chip coupling،inverse taper،FDTD simulation،silicon photonics

چکیده :

Fiber-to-chip coupling remains a critical problem in photonic integrated circuits (PICs), with mode mismatch between single-mode fibers (~10 μm) and silicon waveguides (~0.5 μm) causing significant insertion losses. This paper presents an automated optimization framework combining Python scripting with Finite-Difference Time-Domain (FDTD) electromagnetic simulation for systematic design of inverse taper structures. Using physics-based analytical models validated against FDTD principles, we conduct a comprehensive analysis of the inverse taper geometry involving tip width (0.10-0.30 μm), taper length (50-400 μm), and taper profile shapes. The Differential Evolution optimization process proposed an optimal design of 0.30 μm tip width and 185.7 μm taper length that achieved -2.79 dB coupling efficiency (52.6%) in contrast to -3.22 dB (47.6%) for the baseline designs; thus, a 0.43 dB enhancement was realized. Parametric analysis reveals that tip width significantly impacts scattering loss, while taper length above 150 μm ensures adiabatic mode evolution. The computational framework enables rapid design iteration, with complete optimization completing in under 5 minutes. The computation framework facilitates fast design iteration, with the whole optimization process finishing in less than 5 minutes. The method developed here is a practical tool for photonics engineers to optimize coupling structures for silicon photonics platforms.