Work

"On effective methods to model antenna properties using data-driven model order reduction "

Year: 2024

Type: conference-paper

Abstract: Computational electromagnetic modeling is an essential part of in today's antenna design procedure.The reliability of computational models has significantly increased over the past few decades.Thus, many of the design steps are moved to the computational domain.This also imposes high demands on the simulated antenna models to enable the assessment of their desired performance.This phenomenon of increasing the model phase of the design procedure can be observed both for small Internet-of-Things (IoT) antennas and for array antennas.Developing detailed models of computationally complex devices can be a time-intensive process.This talk will discuss how effective models can reduce computational time and data storage requirements while maintaining high fidelity.

Source: URSI Atlantic Radio Science Conference

DOI: 10.46620/ursiatrasc24/rprm3216

Authors: Lars Jonsson, Harald Hultin, Lucas Åkerstedt

Institution: KTH Royal Institute of Technology

Language: English

Topic: Model Reduction and Neural Networks

Subfield: Statistical and Nonlinear Physics

Field: Physics and Astronomy

Domain: Physical Sciences

Open Access status: bronze

Funder: VINNOVA

Award: Array antennas with reduced number of RF-transmit/receive modules for increased efficiency