The dissertation studies multilayer semiconductor and dielectric microwave devices. The solved problem focuses on improving multilayer microwave devices and on analysing and predicting the frequency characteristics of propagating elec-tromagnetic waves. The dissertation’s object is multilayer semiconductor and dielectric devices, device characteristic analysis and prediction. The work addresses the following tasks: design and improvement of structures of multilayer semiconductor and dielectric microwave devices; calculation and investigation of phase and attenuation characteristics of electromagnetic waves propagating in multilayer microwave devices; prediction of the frequency response of tunable multilayer semiconductor and dielectric devices. The dissertation includes an introduction, four chapters, general conclusions, lists of used literature and the author’s publications on the dissertation’s topic. The introduction analyses the research problem, discusses the relevance of the work, presents the object of research, the purpose of the work, and sets tasks. It presents the work methodology, the scientific novelty of the work, the practical significance of the results, and the defended statements. The first chapter presents an analysis of the sources on the dissertation’s subject, examining the structures of multilayer semiconductor microwave devices and the materials used in the devices. The research methodology and the work tasks are formed. The second chapter discusses the design of tunable multilayer semiconductor devices and presents an investigation of the frequency characteristics of a mi-crostrip bandpass filter on a semiconductor substrate and of the patch antenna with embedded material. The third chapter investigates how the frequency characteristics of phase and attenuation coefficients of multilayer semiconductor waveguide change with changing semiconductor carrier density, the material of the dielectric layer or pa-rameters of the graphene layer. The fourth chapter investigates the possibilities of predicting the characteristics of multilayer semiconductor and dielectric microwave devices using artificial neural networks. Six scientific articles have been published on the dissertation topic: two in journals and two in conference proceedings included in the CA Web of Science list of scientific journals, and two in peer-reviewed IEEE Xplore conference proceedings. The subject of the dissertation was presented at eight scientific conferences: seven international, four of which in Lithuania, three abroad, and one at a national scientific conference.