An artificial neural network based nanoscale device modeling approach and tfet implemantation

Metal Oxide Semiconductor Field Effect Transistor (MOSFET) is a superior market position deal with telecommunications, analogue microelectronics, low-power portables, optoelectronics or other technological areas. The requirements as suggested from the Semiconductor Industry Association (SIA) roadmap may look at alternative routes to mainstream MOSFET that may become economically viable and physical restrictions in some areas of microelectronics in the longer term. Tunnel Field Effect Transistor (TFET) is one of the most important alternatives for MOSFET in The International Technology Roadmap for Semiconductors (ITRS) on the other hand, a combinatorial device to the MOSFETs. TFET is still in research and development process. While the dimensions of the devices decrease, the quantum effects increase. By decreasing the device size, to obtain more realistic device models, quantum effects must be included. However, adding quantum effects cause calculation burden. A modified version of General Regression Neural Network and Non-Equilibrium Green's Function hybrid modeling approach is proposed for accurate nanoscale device model and it is implemented to TFET. In the range of 55% to 67% simulation time decrease is obtained with respect to existing NEGF formalism based methods. In the future works, radiation effects must be added to the model in order to produce radiation hardened devices.