부산대학교 도서관

Printed antennas have played a key role in the quantum leap of portable electronics and communication technology due to their size, robustness and durability. Soft-computational optimization of printed antennas is an emerging trend in antenna design. These methods help in the rapid development of optimized antennas to meet the demands of fast-changing technologies. The geometrical parameters of an antenna are tuned iteratively using an optimization algorithm to best fit the desired performance of an antenna. The use of high-fidelity full-wave simulation is computationally expensive for such iterative evaluation. Low-fidelity surrogate models are viewed as a solution to this problem. This paper reviews some recent works where analytical antenna models are used as surrogate models for soft-computational optimization. Several traditional and modern methods for the analytical modeling of antennas are also reviewed. The analytical models are broadly classified into five categories—traditional analytical methods, antennas modeled as filters, analytical equivalent circuit models, cascade form of analytical equivalent circuit models and computer-aided design of equivalent circuit models. Traditional analytical methods and analytical equivalent circuit models provide higher insights into the working of the antennas. Modeling antennas as filters often result in higher accuracy but limited insights into how the antenna works. The hybrid approach may be viewed as a balance between the two. The computer-aided approach helps in enhancing the accuracy of equivalent circuit models. Examples from each category are reviewed, and they are evaluated based on their applicability in surrogate models.