Title :
Combined genetic algorithm and neural network technique for transistor modeling
Author_Institution :
Electr. & Comput. Eng. Dept., Univ. of Sharjah, Sharjah, United Arab Emirates
Abstract :
A genetic algorithm neural network (GANN) technique for transistor modeling is presented. The proposed method is based on genetic optimization to extract the extrinsic elements of the equivalent circuit model; while the intrinsic part is modelled using a combined genetic neural network approach. The developed procedure has been applied to 1-mm gate width GaN high electron mobility transistor (HEMT) and validated by small- and large-signal measurements.
Keywords :
III-V semiconductors; electronic engineering computing; equivalent circuits; gallium compounds; genetic algorithms; high electron mobility transistors; integrated circuit measurement; integrated circuit modelling; neural chips; GANN technique; GaN; GaN HEMT; equivalent circuit model; extrinsic elements; genetic algorithm neural network technique; genetic optimization; high electron mobility transistor; intrinsic elements; large-signal measurements; small-signal measurements; transistor modeling; Gallium nitride; HEMTs; Integrated circuit modeling; Load modeling; Logic gates; Mathematical model; Optimization; GaN HEMT; genetic optimization; high power amplifier; large-signal modeling; neural networks;
Conference_Titel :
Communications, Signal Processing, and their Applications (ICCSPA), 2015 International Conference on
Conference_Location :
Sharjah
DOI :
10.1109/ICCSPA.2015.7081300
