Modeling carbon nanotube transistors using neural networks approach

A model for carbon nanotube field-effect transistors (CNTFETs) is developed using neural networks approach. This model accurately predicts the I-V characteristics for different structures of CNTFETs. The model is implemented inside the circuit simulator Eldo using its general user defined model (GUDM) template. To confirm the accuracy of the proposed model, the I-V characteristics are compared to device simulation results. The model is also validated using experimental data for both shottky barrier and conventional CNTFETs. The model shows excellent fitting for both the experimental and device simulation data with average percentage error doesn´t exceed 1%.