Abstract:

The Graphene–based Double Gate Field-Effect Transistors (G-DG FETs) have received great attention in recent years due to their high electrical performance provided for analog and Radio-frequency (RF) nanoelectronic applications. To calculate accurately the drain current and the Figures-of-Merit (FoMs) of the nanoscale G-DG FETs requires the solution of Schrödinger/Poisson equations, assuming the quantum effects are to be fully accounted. However, for nanoelectronic circuit simulation, the 2D numerical solution through the fully self-consistent coupled Schrödinger/Poisson equations is an overkill approach in terms of both complexity and computational time cost. Hence, new approach and simulation tools which can be applied to design and simulate Graphene-based nanoelectronic circuits are required to overcome the limitations imposed by the accuracy and computational time cost. In this paper, we investigate the efficiency of a new approach based on the ANN-based computation (Artificial neural network) to analyze and simulate Graphene-based nanoelectronic circuits. In this context, this work presents the applicability of ANN for the simulation of the voltage amplifier by investigating the impact of the G-DG FET design parameters on the analog and RF performances. The ANN-based model can be easily implemented into commercial circuit simulators like: SPICE, Cadence and Silvaco.

Publisher's Version