Prediction of Chaotic Dynamics for Semiconductor Lasers Based on Deep Photonic Reservoir Computing with Time Shifts

The study of the chaotic dynamics of semiconductor lasers is of great theoretical significance and practical application value. However, it remains a challenging task to accurately predict the chaotic dynamic behavior of semiconductor lasers through numerical simulation. Meanwhile, reservoir computing demonstrates powerful and effective capabilities in capturing and predicting the chaotic dynamics of semiconductors. Nevertheless, reservoir computing architectures of different types have their own limitations. Given the challenges in achieving accurate dynamic prediction of semiconductor lasers as described above, we propose to apply the time shifts algorithm to deep reservoir computing to realize the dynamic prediction of lasers. Through numerical analysis, the effectiveness of the time shifts algorithm in predicting the chaotic time series of lasers is verified.