Title :
A novel method for modeling semiconductor by neuro space mapping
Author :
Armaki, Mahdi Gordi ; Anvarifard, Mohamad Kazem ; Hosseini, Seyed Ebrahim
Author_Institution :
Dept. of Eng., Sabzevar Tarbiat Moallem Univ., Sabzevar, Iran
Abstract :
The drift-diffusion (DD) model is not accurate for simulation of submicrometer semiconductor devices. Using RBF NN, a neuro space mapping is proposed to DD model parameters. The DD model with mapped parameters can produce accurate simulation results of the hydrodynamic model. Simulations of a n-i-n diode confirm the ability of the proposed method for submicron device simulation.
Keywords :
electronic engineering computing; neural nets; radial basis function networks; semiconductor device models; semiconductor diodes; drift-diffusion model; hydrodynamic model; n-i-n diode; neuro space mapping; radial basis function neural nets; Artificial neural networks; Circuit simulation; Computational modeling; Electrons; High definition video; Hydrodynamics; Neural networks; Poisson equations; Semiconductor devices; Semiconductor diodes; Drift-diffusion; Neuro space mapping; RBF neural network; n-i-n diode;
Conference_Titel :
Emerging Trends in Electronic and Photonic Devices & Systems, 2009. ELECTRO '09. International Conference on
Conference_Location :
Varanasi
Print_ISBN :
978-1-4244-4846-3
DOI :
10.1109/ELECTRO.2009.5441153
