Title :
EM-ANN models for via interconnects in microstrip circuits
Author :
Watson, P. ; Gupta, K.C.
Author_Institution :
Dept. of Electr. & Comput. Eng., Colorado Univ., Boulder, CO, USA
Abstract :
A novel approach for accurate and efficient modeling of MMIC components by using electromagnetically-trained Artificial Neural Network (EM-ANN) software modules is presented. The approach has been verified by developing models for microstrip via and stripline-to-stripline interconnects in multilayer circuits. Implementation of the approach is demonstrated by integrating the EM-ANN models in commercial microwave circuit simulator.
Keywords :
MMIC; integrated circuit interconnections; integrated circuit modelling; microstrip circuits; neural nets; EM-ANN model; MMIC component; electromagnetically-trained artificial neural network software module; microstrip circuit; microwave circuit simulator; multilayer circuit; stripline-to-stripline interconnect; via interconnect; Artificial neural networks; Backpropagation; Design for experiments; Feedforward systems; Gallium arsenide; Integrated circuit interconnections; Microstrip; Multi-layer neural network; Neurons; US Department of Energy;
Conference_Titel :
Microwave Symposium Digest, 1996., IEEE MTT-S International
Conference_Location :
San Francisco, CA, USA
Print_ISBN :
0-7803-3246-6
DOI :
10.1109/MWSYM.1996.512298
