Multi-objective Optimization of Pixelated Frequency Selective Surface radar absorbing composite using Pareto-based Genetic Algorithm

Abstract

This study proposes and experimentally validates a thin laminated pixelated frequency selective surface (FSS) radar absorbing structure (RAS) optimized using a Pareto-based genetic algorithm (GA). The unit cell was discretized into a 20×20 binary grid with octagonal symmetry to promote polarization-insensitive behavior. The binary pixel topology was optimized together with continuous design variables, including pixel length, sheet resistance, and second dielectric-layer thickness. Two absorption-coverage metrics were used in the optimization: X-band coverage as the primary target-band requirement and broadband coverage as the bandwidth-extension metric. The selected design had a pixel length of 1.25 mm, a sheet resistance of 45.6 Ω/sq, and a second dielectric-layer thickness of 2.31 mm, providing simulated absorption above 90% over 8.12–18.00 GHz. The fabricated specimen had a total thickness of 2.56 mm and an areal density of 3.60 kg/m². Free-space Thru–Reflect–Line-calibrated measurements showed absorption above 90% over 8.37–16.90 GHz for 0° polarization and 8.40–16.01 GHz for 90° polarization. The measured minimum reflection losses were −15.12 dB at 10.66 GHz and −15.56 dB at 10.77 GHz for 0° and 90° polarizations, respectively. A refined sheet-resistance sweep and 25-point sheet-resistance mapping indicated that the simulation–measurement discrepancy was mainly associated with sheet-resistance sensitivity and fabrication-induced nonuniformity. These results demonstrate an experimentally validated design framework for thin laminated pixelated FSS RASs and highlight fabrication-related uncertainty as an important factor in printed resistive FSS-based absorbers.