Title :
Screening of variables for the empirical modelling of semiconductor devices
Author :
Sprevak, D. ; Ferguson, R.S. ; Walton, A.J. ; Fallen, N. ; Newsam, M.I.
Author_Institution :
Dept. of Stat. & Oper. Res., Queen´´s Univ., Belfast, UK
fDate :
9/1/1996 12:00:00 AM
Abstract :
A strategy for screening the important effects in a multivariable response function is presented. An optimisation algorithm is developed, which searches for the optimum of the function in the subspace of the important variables, The application of these methods is illustrated with a known nonlinear 23 variable function. The technique is applied to the design of an NMOS process which depends on 35 factors
Keywords :
MOS integrated circuits; electronic engineering computing; integrated circuit modelling; iterative methods; optimisation; probability; search problems; semiconductor device models; semiconductor process modelling; IC process simulation; NMOS process design; Q-Q plots; active main effects; computer simulation; empirical modelling; iterations; multivariable response function; nonlinear 23 variable function; optimisation algorithm; probability plots; quadratic function; screening of variables; semiconductor devices; subspace of important variables;
Journal_Title :
Science, Measurement and Technology, IEE Proceedings -
DOI :
10.1049/ip-smt:19960485
