Multiobject Design of a 3-D-Printed All-Dielectric Lens Antenna by an Automatic Synthesis Algorithm

Multiperformance indices of an
E
-band three-dimensional (3-D)-printed all-dielectric lens antenna are fulfilled by an automatic synthesis algorithm. By hybridizing particle swarm optimization and genetic algorithm optimization, the algorithm possesses the advantages of fast optimization and avoiding of local optimum. Multiobjective subfunctions are defined to realize multiperformance indices, and the objective subfunctions are customized to make the optimization process more efficient. As the all-dielectric lens is optimized, the source horn is included to direct acquire the performance indices (|
S
₁₁
|, axial ratio, and gain), which makes the optimization process efficient and accurate. Moreover, the optimization process is automatic as the structure modeling and sizing are performed automatically. The proposed lens antenna has good
S
₁₁
, axial ratio, and gain performances in the 60–90 GHz band, which shows that the algorithm is a powerful method for multiple goals.