Research On Modeling And Design Based On Volterra Series For Microwave Nonlinear Scattering Function

With the application of high frequency and high power, microwave components worked in the non-linear zone most of time.In this case, the traditional s-parameter theory was no longer suitable for describing the strongly nonlinear characteristic of the components accurately. Based on the research of nonlinear theory, we introduced a new theory that represented of non-linear characteristics-non-linear scattering function. On the basis of research of non-linear scattering function which had been put forward, we improved it and perfected the concept of it. Taking advantage of this characterization method, we could consider the harmonic characteristics of the components and describe the strongly nonlinear characteristic of the components more accurately.In this thesis, we mainly discussed the technologies of modeling and simulation for Schottky diode, and then the design of Schottky doubler. We started with the traditional modeling methods for Schottky diode and the small-signal model had been extracted by using optimization method.Using the non-linear scattering function testing system in the core of the digital oscilloscope, part of the non-linear scattering function of Schottky diode were measured, and after that we extracted the large-signal model. Then the volterra series method was utilized to analyze the nonlinear characteristics of the component models. In this process, the method we used was nonlinear current method which was included by volterra series analysis. Thereafter we conducted a non-linear scattering function modeling based on volterra series analysis and meanwhile compared with the data we measured. Finally we designed the filter and the matching network of the input and output by using the results of nonlinear scattering function modeling, and completed Schottky Doublers' design.In this thesis, whether the small-signal modeling or the large-signal modeling was based on the data we had measured; therefore the models we conducted were much closer to the actual working state of microwave components. In this way, the too idealistic problem or an internally given bias problem, caused by using the simulation data in the modeling process, could be avoided.