A fuzzy-neural approach for the characterisation of the active microwave devices

12th International Crimean Conference on Microwave and Telecommunication Technology, Sevastopol, Ukrayna, 9 - 13 Eylül 2002, ss.114-117

Yayın Türü: Bildiri / Tam Metin Bildiri
Doi Numarası: 10.1109/crmico.2002.1137168
Basıldığı Şehir: Sevastopol
Basıldığı Ülke: Ukrayna
Sayfa Sayıları: ss.114-117
Yıldız Teknik Üniversitesi Adresli: Evet

Artificial Neural Networks are emerging as a powerful technology for RF and microwave characterization, modeling, and design. 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.

Artificial Neural Networks are emerging as a powerful technology for RF and microwave characterization, modeling, and design. Neural modeler helps us to immediately start developing neural models for RFhlicrowave 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.