A fuzzy-neural approach for the characterisation of the active microwave devices
Bildiriler > Tam Metin Bildiri · 2002
Yazarlar: B. Karlik, H. Torpi, Musa ALCI
12th International Conference on Microwave and Telecommunication Technology, CriMiCo 2002, Amerika Birleşik Devletleri, Sevastopol, Crimea, Uluslararası, Davetli Konuşmacı, Tam Metin Bildiri, 2002-2002, ss 114-117
DOI: 10.1109/CRMICO.2002.1137168
Özet
Artificial neural networks are emerging as a powerful technology for RF and microwave characterization, modeling, and design. A neural modeler helps us to immediately start developing neural models for RF/microwave components and circuits and helps to provide neural models for our simulators. In this study, a novel fuzzy neural network structure is used for behavior of an active microwave device. Here, the device is modeled by a black box whose small signal and noise parameters are evaluated through a fuzzy clustering neural network based upon the fitting of both of these parameters.