Temporal pattern generation with tunable repetition rate using semiconductor laser periodic dynamics.

Area of Science:

• Optoelectronics
• Nonlinear Dynamics
• Signal Processing

Background:

• Semiconductor lasers are crucial for optical communications and signal generation.
• Optical feedback can induce complex dynamics in lasers, offering potential for novel waveform generation.
• Time delay reservoir computing requires diverse and repeatable input signals.

Purpose of the Study:

• To investigate the generation of repetitive temporal patterns with enhanced intra-pattern sample diversity using a long-cavity semiconductor laser with optical feedback.
• To demonstrate stable and continuous generation of periodic waveforms with tunable repetition rates.
• To adapt these generated patterns for applications in time delay reservoir computing.

Main Methods:

• Utilizing a semiconductor laser with a 2-m external cavity and controlled optical feedback.
• Experimentally demonstrating limit cycle dynamics with multiple frequency harmonics.
• Applying high-pass filtering to adjust harmonic power and achieve distinct temporal patterns.

Main Results:

• Stable and continuous generation of periodic waveforms with tunable repetition rates from 3.28 to 4.21 GHz.
• Achieved enhanced intra-pattern sample diversity in generated temporal patterns.
• Demonstrated distinct and highly consistent temporal patterns through harmonic adjustment.

Conclusions:

• A simple long-cavity semiconductor laser configuration with optical feedback can act as a physical waveform generator.
• The generated diverse and repeatable signal patterns are suitable for the masking stage of time delay reservoir computing.
• This approach offers a promising method for generating complex temporal dynamics for advanced computing architectures.