Feature Extraction from Reconfigurable Antenna

Bai, Yuming (2022). Feature Extraction from Reconfigurable Antenna. University of Birmingham. Ph.D.

Preview

Bai2022PhD.pdf
Text - Accepted Version
Available under License All rights reserved.
Download (21MB) | Preview

Abstract

The work in this thesis concerns generic surrogate modelling techniques for a reconfigurable antenna based on any mechanisms. To make sure of the universality of the generic surrogate model, several new perspectives: related to field operating and circuit operating; feature extraction from the antenna; and reconfigurable antenna, are proposed to be a foundation and to provide guidance for the subsequent theory and applications. Additionally, the mathematical derivation of non-linear function fitting and the corresponding equivalent circuit/network is presented as the basic theory. Furthermore, three applications to corresponding antennas are presented to demonstrate the feasibility and effectiveness of the surrogate modelling techniques. Specifically, the first application contains two traditional antennas. They are a chassis antenna with two resonators and a band-notched ultra-wideband pyramidal monopole
antenna. The second application to a frequency reconfigurable UWB antenna with a tunable notched-band is to obtain a data-driven surrogate model. The third application to the same reconfigurable UWB antenna is to obtain a physics-based surrogate model. In these applications, the surrogate modelling approach has many advantages. The approach is reliable and efficient. It has the flexibility for widespread use in many complicated scenarios, because of its elastic order and its adaptable weighting factor. It has fewer extracted parameters with better precision.

These surrogate modelling techniques could be applied to integrate the reconfigurable antenna into a communication system. Furthermore, the approach could also be helpful in the microwave area auto design. Primarily, it could combine artificial intelligence (AI) algorithms to realize future wireless communication such as Smart Antenna, Software Defined Radio and Cognitive Radio.

The novelty and contributions are summarized as follows: (1) This work proposes a completed systematic approach, including fundamental principles, rigorous mathematical derivation and feasible application verification. (2) The approach has the property of generalizability to cover all frequency reconfigurable antennas. (3) As a post-processing approach, it can convert the discrete data of CAD simulation and VNA measurement to a surrogate model with analytical function and equivalent circuit. (4) Some traditional antennas and reconfigurable antennas are taken as examples to demonstrate that the approach is feasible, effective, and precise. (5)
The approach has the flexible ability to adapt to strict requirements and complicated scenarios.

Type of Work:

Thesis (Doctorates > Ph.D.)

Award Type:

Doctorates > Ph.D.

Supervisor(s):