Robust Parameter Tuning of Antenna Structures by Means of Design Specification Adaptation

Parameter tuning through numerical optimization has become instrumental in the design of high-performance antenna systems. Yet, practical optimization faces several major challenges, including high cost of massive evaluations of antenna characteristics, normally involving full-wave electromagnetic (EM) analysis, large numbers of adjustable variables, the shortage of reasonable initial solutions in the case of topologically complex structures, and multimodality of the objective function landscapes. The last two reasons foster the use of global search routines, which might be prohibitively expensive. This article proposes a novel procedure for efficient and reliable antenna optimization by means of design specification adaptation. Using appropriate selection criteria and predictions from a linear expansion model, the ability of improving the current design is verified, and—in the case of a negative outcome—performance requirements are temporarily relaxed. The adjustment is carried out prior to each iteration of the optimization process with the specifications eventually converging to their original levels. The algorithm is intended to improve the efficacy of local optimizers if the target operating frequencies of the structure at hand are severely misaligned with respect to those at the available initial design. This eliminates the need for global routines and greatly enhances the robustness of the search process, as demonstrated using several antenna structures optimized under different scenarios.

I. Introduction

Simulation-driven parameter tuning has become ubiquitous in the design of contemporary antenna structures. As performance specifications become increasingly stringent, which is in a large part dictated by the needs of the emerging application areas (5G communications [1], medical imaging [2], internet of things [3], or wearable devices [4]), their fulfillment fosters the development of topologically complex structures described by many parameters that require meticulous tuning. At the same time, implementation of functionalities such as multiband operation [5], circular polarization [6], band notches [7], or spatial diversity (e.g., MIMO technology [8]) requires handling of various antenna characteristics (axial ratio [9], isolation [10], gain [11]), as well as constraints, for example, in the case of compact realizations [12]. Controlling multiple parameters and performance figures through numerical optimization is a challenging endeavor, yet imperative to render high-performance designs. The primary bottleneck is high cost of repetitive electromagnetic (EM) analyses entailed by both local and global search procedures. To alleviate this difficulty, a number of techniques have been devised, including algorithmic approaches (adjoint sensitivities [13], sparse Jacobian updates [14]), surrogate-based methods involving both physics-based [15], [16] (e.g., space mapping [17], response correction [18], cognition-driven design [19]) and data-driven models [20], [21] (e.g., kriging [22], support vector regression [23]), or machine learning methodologies [24], [25]. Another possibility, partially related to the physics-based approaches, is the use of variable-fidelity simulations, for example, by means of co-kriging [26], multifidelity algorithms [27], or supervised learning [28].