Study of general framework of ferroelectric transistors-based reservoir computing platform
Yiming Cao, Tao Liu, Lican Xiong, Leisheng Jin · Journal of Physics Conference Series · 2024
Abstract In this work, we aim to investigate the inherent correlations between the performance of physical reservoir computing (PRC) and the general physical model of the ferroelectric transistors that is one of the most important candidates for realizing PRC. The physical model of the ferroelectric transistors is constructed based on the Landau-Ginzburg equation. By analyzing the polarisation state of ferroelectrics and the nonlinear dynamics of the model, a PRC that utilizes a single ferroelectric transistor as a physical node is designed, and the typical nonlinear autoregressive moving-average (NARMA) task is employed for verifying the PRC. It is shown that the normalized mean square error (NMSE) can reach 2.8 × 10-3. Finally, the interplay between the different ferroelectric polarisation dynamics and the RC’s performance is discussed by performing numerical simulations on the hyperparameters of the ferroelectric transistor as well as the RC network.