@phdthesis{oai:kitakyu.repo.nii.ac.jp:00000984,
 author = {リュウ, シンユ},
 month = {2022-06-04},
 note = {System modeling is required to deal with the time-varying system dynamics or the experimental data with insufficient information. However, the existing methods cannot construct satisfactory models for rapidly varying systems or severely band-limited signals. This thesis focuses on the new approaches to solve such system modeling problems based on time and frequency-domain information and illustrates their applications in time-varying channel identification and localization system. For the rapid time-varying systems, parameters can be approximated by the cosine series using virtual even periodic functions. Following the orthogonality of the trigonometric functions, the parameter estimation is recursively implemented by estimating the coefficients of each degree of the cosine harmonic term. For the localization system with insufficient frequency components, the spectral characteristics including phase information in frequency domain and the information evaluation in time domain are applied to improve the convergence performance. Numerical simulations demonstrate the effectiveness of the new approaches.},
 school = {北九州市立大学},
 title = {System Modeling and Applications Based on Information in Time and Frequency Domains},
 year = {}
}