.png){kind=icon}
.png){kind=icon}
Return
Enhanced prediction ability for semiconductor laser reservoir computing assisted with silicon-based MRR optical feedback
.png){kind=icon}
PRE
AI
C
L
S
X
M
Y
J
M
A
DOI：10.1007/s11082-026-09071-0.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
High-precision chaotic time-series prediction plays a crucial role in applications such as extreme weather early warning, prevention of large-scale infectious diseases, and structural health monitoring of critical infrastructure. However, achieving accurate predictions for complex chaotic systems remains a significant challenge. In this work, we proposed a novel, performance-enhanced feedback scheme for edge-emitting semiconductor laser (EESL)-based optical reservoir computing (RC) to improve chaotic time-series prediction. The proposed scheme includes a silicon-based micro-ring resonator (MRR) to enrich feedback dynamics. Using two benchmark chaotic time-series [Santa Fe dataset and the tenth-order nonlinear auto-regressive moving average (NARMA10)], we numerically analyzed the prediction performance and underlying mechanisms on the additional physics. Numerical results demonstrate that, compared to conventional optical feedback (COF)-based EESL-RC, the proposed RC system exhibits better computational performance, particularly for complex time-series like NARMA10, which demand higher memory capacity (MC). The enhanced MC achieved by our scheme was numerically validated against COF. Furthermore, we investigated the influence of key parameters—including EESL feedback strength, MRR feedback constant, and feedback detuning—on the nonlinear dynamics and computational performance. This study can provide a valuable insight for advancing high-performance photonic neuromorphic computing systems based on RC platforms.
Keywords:
Photonic reservoir computing
Semiconductor laser
Silicon-based micro-ring resonator
Prediction ability
Memory capacity
Journal
IF:
4
Papers:
9.8K
Citations:
1.8W
Organization
Cited Papers