Title:	Cloud-based machine learning techniques implemented by microsoft azure for designing power amplifiers
Authors:	Jamshidi, Mohammad Roshani, Saeed Talla, Jakub Sharifi-Atashgah, Maryam S. Roshani, Sobhan Peroutka, Zdeněk
Citation:	JAMSHIDI, M. ROSHANI, S. TALLA, J. SHARIFI-ATASHGAH, MS. ROSHANI, S. PEROUTKA, Z. Cloud-based machine learning techniques implemented by microsoft azure for designing power amplifiers. In Proceedings of 2021 IEEE 12th Annual Ubiquitous Computing, Electronics & Mobile Communication Conference (IEEE UEMCON). Piscaway: IEEE, 2021. s. 0041-0044. ISBN: 978-1-66540-690-1
Issue Date:	2021
Publisher:	IEEE
Document type:	konferenční příspěvek ConferenceObject
URI:	http://hdl.handle.net/11025/47093
ISBN:	978-1-66540-690-1
Keywords in different language:	artificial intelligence;cloud computing;power amplifiers;machine learning;deep learning;Microsoft Azure;power electronics
Abstract in different language:	Designing power amplifiers based on the demanded power and frequency is one of the challenging processes of circuits design in electrical engineering. This is best understood when it comes to thermal noises and other unwanted agents. This is why the application of cloud-based methods can be beneficial to save time and money for designing such complex systems. In this paper, several machine learning (ML) approaches have been used to design a class E amplifier. In this regard, the proposed methods, which are implemented via Microsoft Azure, are used to model and predict the circuit element values of the class E amplifier. In order to reach a reliable design, some important unwanted factors such as nonlinear parasitic elements of the transistor are considered. The results demonstrated that not only can the proposed could-based techniques estimate such elements accurately, but also working with such tools are really easy.
Rights:	Plný text je přístupný v rámci univerzity přihlášeným uživatelům. © IEEE
Appears in Collections:	Konferenční příspěvky / Conference Papers (KEV) Konferenční příspěvky / Conference papers (RICE) OBD

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.