Automated control algorithms for silicon photonic polarization receiver

Area of Science:

• Photonics
• Optical Communications
• Control Systems

Background:

• Optical fiber communication systems are susceptible to polarization fluctuations.
• Maintaining a stable polarization state is crucial for high-speed optical signal reception.
• Silicon photonic devices offer miniaturization and integration potential for optical receivers.

Purpose of the Study:

• To develop and evaluate automated control algorithms for silicon photonic polarization receivers.
• To enable real-time adaptation and stabilization of arbitrary optical fiber polarization states.
• To compare the performance of different gradient descent and optimization-based control strategies.

Main Methods:

• Implementation of greedy linear descent, basic gradient descent, two-point step size gradient descent, and two-stage optimization algorithms.
• Utilizing an active feedback loop for polarization control.
• Testing with a 10 Gb/s on-off keying signal in a silicon photonic polarization receiver.
• Conducting long-duration experiments to assess continuous polarization management.

Main Results:

• Successful automated adaptation of time-varying polarization states to the transverse-electric (TE) mode.
• Demonstrated real-time adaptation for a 10 Gb/s signal.
• Comparison of algorithms based on iteration count and output response.
• Long-duration experiments confirmed continuous polarization management capabilities.

Conclusions:

• The developed automated control algorithms effectively manage polarization in silicon photonic receivers.
• Gradient descent and two-stage optimization methods provide robust, real-time polarization stabilization.
• These techniques are vital for reliable high-speed optical communication systems.