An extraction technique for small signal intrinsic parameters of HEMTs based on artificial neural networks

摘要: Abstract This paper presents a fast and accurate procedure for extraction of small signal intrinsic parameters AlGaAs/GaAs high electron mobility transistors (HEMTs) using artificial neural network (ANN) techniques. The has been done in wide range frequencies biases at various temperatures. Intrinsic HEMT are acquired its values common-source S-parameters. Two different ANN structures have constructed this work to extract the parameters, multi layer perceptron (MLP) radial basis function (RBF) networks. These two kinds ANNs compared each other terms accuracy, speed memory usage. To validate capability proposed method modeling GaAs HEMTs, data modeled S-parameters 200 μm gate width 0.25 μm very good agreement between them is achieved up 30 GHz. effect bias, temperature frequency conditions on extracted investigated, obtained results match theoretical expectations. model can be inserted computer-aided design (CAD) tools order an design, simulation optimization microwave circuits including HEMTs.