Novi pristupi u razvoju talasnog modela šuma mikrotalasnih tranzistora
Đorđević, Vladica N.
Faculty:Универзитет у Нишу, Електронски факултет
MetadataShow full item record
- Истраживање и развој решења за побољшање перформанси бежичних комуникационих система у микроталасном и милиметарском опсегу фреквенција (MPNTR-TR 32052)
The microwave transistor noise wave model is defined by the parameters called the noise wave temperatures. These temperatures are determined based on the measured transistor noise parameters, mostly by using optimization procedures in the microwave circuit simulators. The main disadvantage of the optimization procedures is that they are usually time-consuming. Therefore, the aim of the research presented in this dissertation is to find the new extraction methods for more efficient determination of the noise wave temperatures. New extraction methods both for the indirect and direct determination of the noise wave temperatures are developed. In the first case, the noise wave temperatures are calculated on the basis of the extracted noise parameters of the transistor intrinsic circuit, by using appropriate equations. The noise parameters of the transistor intrinsic circuit are extracted on the basis of the measured transistor noise parameters, by applying one of the following three de-embedding
procedures: analytical de-embedding procedure, de-embedding procedure within the microwave circuit simulator or de-embedding procedure based on artificial neural networks. In the second case, the determination of the noise wave temperatures is performed directly, on the basis of the measured transistor noise parameters. For the purpose of the direct determination of the noise wave temperatures, four extraction methods based on artificial neural networks, as well as semi-analytical extraction method based on polynomials are developed. In order to provide their validation, all proposed extraction methods are used for the modeling of noise characteristics of the specific GaAs HEMT component in the wide range of operating temperatures. In addition, the noise wave model application for the noise parameter modeling in the case of GaN HEMT, illuminated GaAs HEMT, and three different GaAs HEMTs with the scaled gate width is presented in the dissertation. In all cases, some of the extraction methods for the indirect determination of the noise wave temperatures is used. Hence, the reliability of the noise wave model is proved in terms of the new generation of components, the components exposed to various light conditions as well as the same class components with different gate widths.View More
Keywords:Mikrotalasni tranzistori; Artificial neural networks; Modelovanje šuma; Parametri šuma; Talasni model šuma; Talasne temperature šuma; Veštačke neuronske mreže; Microwave transistors; Noise modeling; Noise parameters; Noise wave model; Noise wave temperatures