Abstract
Abstract
Metasurfaces enable precise manipulation of electromagnetic (EM) waves for applications such as beam steering, sensing, and stealth technology. However, inverse design of metasurfaces remains challenging owing to the computational burden of iterative full-wave simulation-driven optimization, the difficulty of achieving controllable generation under continuous EM constraints, and the inherently non-unique nature of the inverse mapping. To address these challenges, this paper presents a generative inverse design framework for controllable and physically consistent metasurface synthesis under continuous spectral constraints. The proposed approach employs a progressively growing Wasserstein generative adversarial network with gradient penalty (WGAN-GP) integrated with feature-wise linear modulation (FiLM)-based conditioning for stable propagation of continuous spectral and fabrication constraints. EM consistency is embedded directly into the generative learning process through a surrogate-assisted spectral alignment loss, enabling physics-constrained generation during training. Further, a determinantal point process (DPP)-based diversity regularization strategy is incorporated to generate geometrically diverse yet spectrally consistent realizations for the same target response. The effectiveness of the proposed framework is demonstrated through the generation of practically realizable metasurface absorbers exhibiting diverse reflection characteristics in the frequency range of 2–18 GHz. EM simulations validate that the generated designs meet the target specifications with high accuracy. The final proposed framework achieved an average mean squared error of 0.0052, accumulated average error of 0.0343, diversity score of 0.8730, band alignment accuracy of 0.8533, and a valid EM design generation rate of 89.57%, clearly demonstrating its capability to generate highly accurate, diverse, electromagnetically consistent and fabrication realizable metasurface configurations.