Litcius/Paper detail

Frequency-Based Regularization for Improved Reliability Optimization of Antenna Structures

Sławomir Kozieł, Anna Pietrenko‐Dabrowska, Muath Al‐Hasan

2020IEEE Transactions on Antennas and Propagation20 citationsDOI

Abstract

This communication proposes a modified formulation of antenna parameter tuning problem. The main ingredient of the presented approach is a frequency-based regularization. It allows for smoothening the functional landscape of the assumed cost function, defined to encode the prescribed design specifications. The regularization is implemented as a special penalty term complementing the primary objective and enforcing the alignment of the antenna responses with the target operating frequency (or frequencies for multiband antennas). The result is an improved reliability and reduced cost of the optimization process, both highly desirable from the point of view of the efficacy of electromagnetic (EM)-driven design procedures. Furthermore, regularization makes the use of local routines sufficient even in situations where the global search is otherwise imperative (e.g., due to poor initial design). Our methodology is demonstrated using three microstrip antennas. The superior performance of the approach, both in terms of design reliability and in terms of the computational cost of the optimization process, is corroborated by comparisons with a conventional formulation.

Topics & Concepts

Regularization (linguistics)Computer scienceMinificationMathematical optimizationReliability (semiconductor)Microstrip antennaAntenna (radio)MathematicsTelecommunicationsPhysicsArtificial intelligenceQuantum mechanicsPower (physics)Antenna Design and OptimizationEnergy Harvesting in Wireless NetworksAdvanced Multi-Objective Optimization Algorithms
Frequency-Based Regularization for Improved Reliability Optimization of Antenna Structures | Litcius