Optimal array design of deficient coding metasurfaces for terahertz beam steering via phase-gradient approaching

• Published in: Optics express Vol. 33; no. 9; p. 20095
• Main Authors: Li, Yueting, Lan, Feng, Hu, Tianyu, Nie, Xiaolei, Yang, Jiayao, Wang, Luyang, Song, Tianyang, Yang, Munan, Mazumder, Pinaki, Zhang, Yaxin, Yang, Ziqiang
• Format: Journal Article
• Language: English
• Published: United States 05.05.2025
• ISSN: 1094-4087; 1094-4087
• DOI: 10.1364/OE.558331

Confronted with the low reconfigurability of switches and materials with continuously tunable refractive indices, terahertz active coding metasurfaces suffer from narrow phase shift ranges and low quantization precision. To address the consequent terahertz beam-steering deterioration, an array-level optimization framework from the principle of phase-coding modulation and genetic algorithm (GA) has evolved, enabling enhanced beam steering on deficient 2-bit coding metasurfaces. Unlike traditional exhaustive iterative optimization methods, a near-optimal initial population from a phase-gradient approaching algorithm combined with objective directivity functions significantly improves the beam steering performance in fewer iterations. A passive, 2-bit coding metasurface is designed as a proof-of-concept instead of the reconfigurable coding metasurface. The simulation and experimental results under normal and oblique incidences show that, compared to fractional coding(F code), the phase-gradient approaching (PGA) method with a 150° phase modulation range increases the main lobe amplitude by an average of 2.29 dB (25.92%) and reduces the sidelobe level (SLL) by an average of 2.52 dB (26.45%) within the range of 21°–60° at 220 GHz. Moreover, the PGA-initiated GA optimization increases the computational efficiency by reducing the total iterations by 61.08% compared to F code optimization. The proposed method offers a feasible array optimization solution for non-ideal coding metasurfaces, with potential applications in terahertz wireless communication, computational imaging, and compressive sensing.