The influence of timescales and data injection schemes for reservoir computing using spin-VCSELs

Abstract

Reservoir computing with photonic systems promises fast and energy efficient computations. Vertical emitting semiconductor lasers with two spin-polarized charge-carrier populations (spin-VCSEL), are good candidates for high-speed reservoir computing. With our work, we highlight the role of the internal dynamic coupling on the prediction performance. We present numerical evidence for the critical impact of different data injection schemes and internal timescales. A central finding is that the internal dynamics of all dynamical degrees of freedom can only be utilized if an appropriate perturbation via the input is chosen as data injection scheme. If the data is encoded via an optical phase difference, the internal spin-polarized carrier dynamics is not addressed but instead a faster data injection rate is possible. We find strong correlations of the prediction performance with the system response time and the underlying delay-induced bifurcation structure, which allows to transfer the results to other physical reservoir computing systems. The authors numerically investigate the reservoir computing performance of vertical emitting two-mode semiconductor lasers and show the crucial impact of dynamic coupling, injection schemes and system timescales. A central finding is that high dimensional internal dynamics can only be utilized if an appropriate perturbation via the input is chosen.