High precision wavelength estimation method for integrated optics.

Area of Science:

• Photonics and Optical Engineering
• Machine Learning Applications
• Spectroscopy

Background:

• Accurate optical wavelength measurement is crucial for various photonic applications.
• Existing methods often involve complex setups or limited precision.
• Need for simplified and integrated optical sensing solutions.

Purpose of the Study:

• To introduce a novel, simple, and highly precise optical wavelength measurement approach.
• To demonstrate the system's effectiveness using a tunable waveguide micro ring resonator and a photodiode.
• To enable optical integration through a simplified system design.

Main Methods:

• Calibration using known wavelengths and resonator tuning voltages.
• Modeling the spectral sensitivity function with a neural network.
• Estimating unknown wavelengths via photodiode signals, neural networks, and optimization algorithms.

Main Results:

• Achieved high spectral precision of approximately 8 pm (5 × 10⁻⁶ at 1550 nm) across a 4 nm range.
• Attained even higher precision of 5 pm (3 × 10⁻⁶) in a narrower band (1550.3–1550.8 nm).
• Demonstrated a five-fold improvement in precision compared to simple data lookup.

Conclusions:

• The proposed method offers a significant simplification of optical systems for wavelength measurement.
• The approach facilitates optical integration, paving the way for compact devices.
• The methodology is broadly applicable to adjustable wavelength monitoring devices.