Modeling of Ballistic Carbon Nanotube Transistors by Neural Space Mapping

In this paper we present a neural space mapping (NSM) SPICE-compatible modeling technique for carbon nanotubes in their ballistic limit. We modified conventional MOSFET equations and used NSM concept in order to correct these equations in the manner that to be usable to carbon nanotube transistors. We used Fettoy model (which is applicable to MOSCNT structures) to compare accuracy. We have shown that our model has reasonable accuracy at different biases and physical parameters.