Title :
Neural Networks for the Design and Reverse Engineering of BJTs
Author :
Ferguson, Ryan ; Roulston, David J.
Author_Institution :
Department of Electrical and Computer Engineering, University of Waterloo, Waterloo, Ontario, Canada N2L 3G1
Abstract :
We have developed a neural network based software tool, REED (Rapid Engineering of Electron Devices) that maps an electrical description of a bipolar transistor to its physical structure, as described by mask geometries and a doping profile.
Keywords :
Backpropagation algorithms; Bipolar transistors; Design engineering; Doping profiles; Electrons; Neural networks; Reverse engineering; SPICE; Semiconductor process modeling; Software tools;
Conference_Titel :
Solid State Device Research Conference, 1996. ESSDERC '96. Proceedings of the 26th European
Conference_Location :
Bologna, Italy
