Summary
International Symposium on Nonlinear Theory and its Applications
2017
Session Number:A3L-C
Session:
Number:A3L-C-4
Information Processing Based on an External-Cavity Semiconductor Laser with Optical Feedback-Phase Modulation
Kazutaka Kanno, Chihiro Sugano, Kousuke Takano, Atsushi Uchida, Masatoshi Bunsen,
pp.264-267
Publication Date:2017/12/4
Online ISSN:2188-5079
Summary:
Reservoir computing (RC) based on a semiconductor laser with time-delayed optical feedback is numerically demonstrated in this study. RC is a machine learning paradigm based on information processing in the human brain. In our system, input information is injected into the laser via the phase of its optical feedback and any other lasers are not used for information input. We demonstrated a chaotic time-series prediction by our RC system and investigated the dependence of the prediction error on the feedback strength, the laser injection current, and a node interval by which a feedback loop is temporally divided for considering virtual nodes. The numerical analysis showed that the prediction error reduces as the injection current is increased when the feedback strength and the node interval are adjusted. Because signal to noise ratio is improved at a large value of the injection current.