A Dual-Polarized 5G Base Station Antenna Using Machine-Learning Based Optimization Method

  • Hua, Q. (Speaker)
  • Yi Huang (Contributor to Paper or Presentation)
  • Chaoyun Song (Contributor to Paper or Presentation)
  • Tianyuan Jia (Contributor to Paper or Presentation)
  • Xu Zhu (Contributor to Paper or Presentation)

Activity: Talk or presentation typesOral presentation


A broadband dual-polarized 5G base station antenna using a machine-learning based optimization method is presented in this paper, which covers 3.3 – 5.0 GHz and has a compact size with an overall dimension of 60 × 60 × 18 mm3. The antenna includes two double-oval-shaped dipoles, two Г-shaped feeding lines and one reflector. The double-oval-shaped dipoles and Г-shaped feeding lines can generate a broadband performance and high port-to-port isolation. The training cost reduced surrogate model-assisted hybrid differential evolution for complex antenna synthesis (TR-SADEA) is employed to optimize the overall antenna performance. An optimization procedure is also introduced in the paper. The proposed antenna has a high port-to-port isolation (better than 20 dB), a stable radiation pattern, and a flat gain (about 7.3 dBi). Meanwhile, the half-power beamwidth of the proposed antenna is within 65° ± 5°, which meets the industry requirement. Therefore, the proposed antenna is a good candidate for 5G base station applications.
Period17 Mar 2020
Event title14th European Conference on Antennas and Propagation
Event typeConference
Conference number14
LocationCopenhagen, DenmarkShow on map
Degree of RecognitionInternational