FOURIER METHOD MODELING OF SEMICONDUCTOR-DEVICES

被引:3
作者
AXELRAD, V
机构
[1] Lehrstuhl für Integrerte Schaltungen, Technische Universität München
关键词
D O I
10.1109/43.62760
中图分类号
TP3 [计算技术、计算机技术];
学科分类号
0812 ;
摘要
A novel high-order approach to numerical modeling of semiconductor devices is presented. The new method is a combination of the classical Fourier-series Galerkin procedure, a special matrix calculus, and fast numerical pseudospectral techniques. The proposed algorithm renders the exact solution (machine precision) of the semiconductor equations in closed form of a trigonometric polynomial. The condition number of the diagonally dominant discrete equations is near unity. As a consequence, a highly accurate solution is achieved at moderate computer costs. The method has been implemented for one- and two-dimensional device models. Properties of the new procedure are demonstrated on examples. © 1990 IEEE
引用
收藏
页码:1225 / 1237
页数:13
相关论文
共 21 条
[1]  
ASCHER U, 1986, UBC8612 COMP SCI TEC
[2]  
Axelrad V., 1989, International Journal of Numerical Modelling: Electronic Networks, Devices and Fields, V2, P31, DOI 10.1002/jnm.1660020104
[3]  
AXELRAD V, 1987, THESIS TU MUNCHEN
[4]  
AXELRAD V, 1989, P INT S VLSI TECHNOL
[5]  
AXELRAD V, 1987, THEORY FLEXIBLE SHEL
[6]  
AXELRAD V, 1989, 6 P NASECODE DUBL
[8]  
Dongarra J. J., 1979, LINPACK USERS GUIDE
[9]  
Gilbarg D., 1983, ELLIPTIC PARTIAL DIF
[10]  
Korn GA, 1968, MATH HDB SCI ENG, V2nd