TY - JOUR
T1 - A two-step optimization algorithm for enhanced sidelobe suppression ratio and robust beamforming in optical phased arrays
AU - Lin, Jian
AU - Liang, Jiachang
AU - Liu, Xiaoping
AU - Huang, Mingda
AU - Wang, Feihu
AU - Wang, Yuanfeng
AU - Hu, Xiaopeng
AU - Wu, Xiaoshan
AU - Cai, Hongling
AU - Qin, Qi
PY - 2026/10
Y1 - 2026/10
N2 - As a compact integrated alternative to traditional microwave phased arrays, optical phased arrays (OPAs) have demonstrated great potential across a wide range of applications. Beamforming performance, particularly the sidelobe suppression ratio (SLSR), is crucial for signal quality and interference immunity. To enhance SLSR, we propose a hybrid two-step optimization algorithm that combines adaptive moment estimation (Adam) and particle swarm optimization (PSO). Using a composite variable splitter (VS), we achieve average SLSRs over a ±70° scanning range >20.14 dB for Bayliss difference patterns and >22.23 dB for Dolph-Chebyshev patterns. For asymmetric multi-beam patterns, the achieved average SLSRs are 20.14 dB, 19.10 dB, and 18.11 dB for the four-, six-, and eight-beam patterns, respectively. Robustness to VS splitting-ratio (SR) imbalance is further demonstrated using a traditional VS with SR constrained to ±13.80 dB, where the average SLSRs remain >19.07 dB and >20.87 dB for the Bayliss difference and Dolph-Chebyshev patterns and 18.42 dB, 17.12 dB, and 16.83 dB for the four-, six-, and eight-beam patterns, respectively. These results show that the proposed algorithm improves OPA beamforming quality while relaxing the VS precision required to support a wide SR tuning range, facilitating chip-level integration of large-scale, high-performance OPAs. © 2026 Elsevier B.V.
AB - As a compact integrated alternative to traditional microwave phased arrays, optical phased arrays (OPAs) have demonstrated great potential across a wide range of applications. Beamforming performance, particularly the sidelobe suppression ratio (SLSR), is crucial for signal quality and interference immunity. To enhance SLSR, we propose a hybrid two-step optimization algorithm that combines adaptive moment estimation (Adam) and particle swarm optimization (PSO). Using a composite variable splitter (VS), we achieve average SLSRs over a ±70° scanning range >20.14 dB for Bayliss difference patterns and >22.23 dB for Dolph-Chebyshev patterns. For asymmetric multi-beam patterns, the achieved average SLSRs are 20.14 dB, 19.10 dB, and 18.11 dB for the four-, six-, and eight-beam patterns, respectively. Robustness to VS splitting-ratio (SR) imbalance is further demonstrated using a traditional VS with SR constrained to ±13.80 dB, where the average SLSRs remain >19.07 dB and >20.87 dB for the Bayliss difference and Dolph-Chebyshev patterns and 18.42 dB, 17.12 dB, and 16.83 dB for the four-, six-, and eight-beam patterns, respectively. These results show that the proposed algorithm improves OPA beamforming quality while relaxing the VS precision required to support a wide SR tuning range, facilitating chip-level integration of large-scale, high-performance OPAs. © 2026 Elsevier B.V.
UR - https://www.scopus.com/pages/publications/105039122376
UR - https://www.scopus.com/pages/publications/105039122376#tab=impact
U2 - 10.1016/j.optcom.2026.133379
DO - 10.1016/j.optcom.2026.133379
M3 - RGC 21 - Publication in refereed journal
SN - 0030-4018
VL - 616
JO - Optics Communications
JF - Optics Communications
M1 - 133379
ER -