Prior knowledge input neural network method for GFET description
For circuit design, various compact models for graphene field-effect transistors (GFETs) have been developed. However, the consistency between them is poor, since study on the mechanism of operation of GFETs is still immature and the models were derived based on different understandings. Herein, we propose another approach for circuit-level description of GFETs based on a prior knowledge input neural network modeling method. By virtue of the neural network’s learning ability, it can accurately describe different GFETs without an exact description of their mechanism of operation. Basic knowledge on GFETs and an adaptive genetic algorithm are employed to improve the precision of the neural network, resulting in performance significantly exceeding that of a traditional, multilayer perceptron network. The universality of the method is verified by detailed tests using two different datasets. Its applicability for circuit design is demonstrated by relevant circuit simulations in a Verilog-A implementation.
KeywordsGenetic algorithm Graphene field-effect transistors Neural network Prior knowledge Verilog-A
This work was supported by the National Natural Science Foundation of China (61204096, 61404094, and 61574102), the Fundamental Research Fund for the Central Universities, Wuhan University (2042014kf0238 and 2042015kf0174), the China Postdoctoral Science Foundation (2012T50688), the Natural Science Foundation of Hubei Province, China (2014CFB694), and the Science Foundation of Jiangsu Province, China (BK20141218).
- 10.Wang, H., Nezich, D., Kong, J., Palacios, T.: Graphene frequency multipliers. IEEE Electr. Device Lett. 30(5), 547–549 (2009)Google Scholar
- 11.Lin, Y.M., Valdes-Garcia, A., Han, S.J., Farmer, D.B., Meric, I., Sun, Y., Wu, Y., Dimitrakopoulos, C., Grill, A., Avouris, P., Jenkins, K.A.: Wafer-scale graphene integrated circuit. Science 332(6035), 1294–1297 (2011)Google Scholar
- 12.Han, S.J., Garcia, A.V., Oida, S., Jenkins, K.A., Haensch, W.: Graphene radio frequency receiver integrated circuit. Nat. Commun. 5 (2014)Google Scholar
- 13.Nguyen, V.H., Bournel, A., Chassat, C., Dollfus, P.: Quantum transport of Dirac fermions in graphene field effect transistors. In: International Conference on Simulation of Semiconductor Processes and Devices (SISPAD), Bologna, pp. 9–12 (2010)Google Scholar
- 16.Henry, M.B., Das, S.: SPICE-compatible compact model for graphene field-effect transistors. In: IEEE International Symposium on Circuits and Systems (ISCAS), Seoul, pp. 2521–2524 (2012)Google Scholar
- 18.Zhang, Q.J., Gupta, K.C.: Neural networks for RF and microwave design (Book+ Neuromodeler Disk). Artech House Inc, Norwood (2000)Google Scholar
- 21.Gao, J., Zhang, L., Xu, J., Zhang, Q.J.: Nonlinear HEMT modeling using artificial neural network technique. In: International Microwave Symposium Digest (IEEE MTT-S), p. 4 (2005)Google Scholar
- 22.Watson, P.M., Gupta, K.C., Mahajan, R.L.: Development of knowledge based artificial neural network models for microwave components. In: International Microwave Symposium Digest (IEEE MTT-S), Baltimore, vol. 1, pp. 9-12 (1998)Google Scholar
- 28.Foresee, F.D., Hagan, M.T.: Gauss-Newton approximation to Bayesian regularization. Joint Conf. Neural Netw. 1997, 1930–1935 (1997)Google Scholar
- 29.Ngugen, D., Widraw, B.: Improving the learning speed of 2-layer neural networks by choosing initial values of the adaptive weights. In: Procceeding of the International Joint Conference on Neural Networks, San Diego, 3, pp. 21–26 (1990)Google Scholar