EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Integrated Design Framework for Scalable Series-Fed Microstrip Patch Antenna Arrays With Machine Learning Assistance

R. Priya, R. V. Sanjika Devi, Parul Mathur, Robin Augustine, Dhanesh G. Kurup

IEEE Access · 2026 · 14 · 48478-48486

DOI: 10.1109/access.2026.3678079 ↗

Kurz gefasst

Fachartikel in der Zeitschrift IEEE Access, 2026. Verfasst von R. Priya und 4 weiteren Autor:innen. Die Arbeit nennt 24 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

This work presents an integrated design framework for compact series-fed microstrip patch antenna arrays (SF-MSPAAs) enabling scalable sub-arrays for planar and MIMO configurations. In the proposed framework, an artificial neural network (ANN) based surrogate model is utilized as a computational accelerator to eliminate repeated full-wave simulations, while the subsequent array design is carried out using deterministic, physics-based electromagnetic design procedures. Phase equalization is achieved through S-parameter based optimization of feed-line lengths using sequential nonlinear programming (SNLP), resulting in quasi-uniform inter-element spacing without additional phase compensation structures. The desired sidelobe level (SLL) is realized via width-scaled Taylor amplitude tapering, and robust input matching is ensured using an adaptive two-section microstrip impedance transformer. Unlike conventional approaches that primarily optimize one or few performance metrics aggressively, for example, maximum sidelobe suppression, the proposed framework provides specification-driven control over key parameters such as operating frequency, SLL, gain and array footprint. This allows designs to precisely meet application-specific requirements within a structured and computationally efficient workflow. A five element prototype is designed using the proposed framework to meet an application requirement of $SLL\leq {-15}$ dB and a gain of $\geq 10$ dBi operating at $6.4{\,}\mathrm {GHz}$ . The designed prototype is fabricated and experimentally validated. The measured results demonstrated a gain of $\sim \!12{\,}\mathrm {dBi}$ , 75% radiation efficiency, 1.6% bandwidth, ${-18}{\,}\mathrm {dB}$ SLL and return loss better than $35{\,}\mathrm {dB}$ across the bandwidth, showing close agreement with simulations.

Zitieren

APA

Priya, R., Devi, R. V. S., Mathur, P., Augustine, R., & Kurup, D. G. (2026). Integrated Design Framework for Scalable Series-Fed Microstrip Patch Antenna Arrays With Machine Learning Assistance. IEEE Access, 14, 48478-48486. https://doi.org/10.1109/access.2026.3678079

Vancouver

Priya R, Devi RVS, Mathur P, Augustine R, Kurup DG. Integrated Design Framework for Scalable Series-Fed Microstrip Patch Antenna Arrays With Machine Learning Assistance. IEEE Access. 2026;14:48478-48486. doi:10.1109/access.2026.3678079

BibTeX
@article{Priya2026,
  author = {Priya, R. and Devi, R. V. Sanjika and Mathur, Parul and Augustine, Robin and Kurup, Dhanesh G.},
  title = {Integrated Design Framework for Scalable Series-Fed Microstrip Patch Antenna Arrays With Machine Learning Assistance},
  journal = {IEEE Access},
  year = {2026},
  volume = {14},
  pages = {48478--48486},
  doi = {10.1109/access.2026.3678079}
}
Kostenlos starten →