{
  "id": 2066626,
  "title": "Machine learning-based design and optimization of a cross-shaped circularly polarized multiband patch MIMO antenna for 6G sub-THz communication systems",
  "url": "https://urgent.news/2026/08/20/machine-learning-based-design-and-optimization-of-a-cross-shaped",
  "topic": "ai",
  "section": "AI",
  "published": "2026-08-20T00:00:00.000Z",
  "source": {
    "name": "Scientific Reports",
    "slug": "scientific-reports",
    "url": "https://www.nature.com/articles/s41598-026-67438-5"
  },
  "original_language": "en",
  "account": "This study introduces a machine learning-based approach to design and optimize a multiband, multiple-input multiple-output (MIMO) patch antenna tailored for sub-terahertz (sub-THz) communication systems. The researchers utilized Genetic Algorithm (GA), Particle Swarm Optimization (PSO), and Differential Evolution (DE) to fine-tune the antenna's characteristics. A comprehensive suite of machine learning models was evaluated, with the Random Forest model standing out due to its exceptional performance, as indicated by high \\(R^{2}\\) values of 98.18%. This model proved superior in predicting the antenna's behavior, balancing prediction accuracy, convergence time, mean absolute error (MAE), and the coefficient of determination (\\(R^{2}\\)). The proposed antenna incorporates an asymmetric cross-shaped slotted patch design on a Rogers RT5880 dielectric substrate, resulting in circular polarization. Operating within the frequency range of 134 GHz, 230 GHz, and 450 GHz, the antenna exhibits bandwidths of 5.98 GHz, 14.64 GHz, and 82 GHz, respectively. The single-element design delivers a gain of 6.78 dBi, while an optimized linear array configuration achieves a peak gain of 13.9 dBi. Notably, the single-element antenna maintains a radiation efficiency of up to 80% and demonstrates high isolation levels below –30 dB across the operating frequencies, all without the need for a defected ground structure. The circular polarization is validated with an axial ratio below 3 dB over the entire frequency range. The machine learning-assisted surrogate optimization process significantly reduces computational time, complexity, and resource demands while preserving high prediction accuracy. This research introduces a novel, multi-objective optimization framework specifically designed for the creation of multiband, circularly polarized MIMO antennas operating in the sub-THz frequency band. The study was conducted at Bahir Dar University and did not receive external funding.",
  "summary": "Scientific Reports, Published online: 20 August 2026; doi:10.1038/s41598-026-67438-5 Machine learning-based design and optimization of a cross-shaped circularly polarized multiband patch MIMO antenna for 6G sub-THz communication systems",
  "key_points": [],
  "editors_take": null,
  "illustration": null,
  "coverage": {
    "outlets": 1,
    "also_reported_by": []
  },
  "ai_generated": true,
  "disclaimer": "Summaries, key points and the editor’s take are written by software from other outlets’ reporting and may contain errors — always check the linked original."
}