EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Reasoning and Acting (ReAct) With Multimodal LLMs: A Framework for Intent Driven 6G Networks

Syed Ali Raza Zaidi, Maryam Hafeez, Mohammed M. H. Qazzaz, Mallik Tatipamula, Kuntal Chowdhury, Moe Z. Win

IEEE Open Journal of the Communications Society · 2026 · 7 · 5463-5481

DOI: 10.1109/ojcoms.2026.3693067 ↗

Kurz gefasst

Fachartikel in der Zeitschrift IEEE Open Journal of the Communications Society, 2026. Verfasst von Syed Ali Raza Zaidi und 5 weiteren Autor:innen. Die Arbeit wurde bisher 2-mal zitiert (Crossref) und nennt 102 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Advanced Connectivity Technology Hub, Communications Hub for Empowering Distributed Cloud Computing Applications and Research (CHEDDAR) through the Engineering…, Academic Grant from Google Inc. for exploiting the GenAI capabilities in VertexAI, Department for Science, Innovation and Technology (DSIT), the U.K. AI Security Institute (AISI), and EPSRC.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

6G wireless networks are poised to realise Self-X capabilities through the native integration of an AI-plane within the network architecture. This AI-plane will enable autonomous orchestration and management driven by user intent. Recent advances in Generative AI (GenAI) introduce powerful capabilities for such intent-based management, particularly through the representation of multimodal data in a shared embedding space. This enables the development of autonomous agents that not only translate intent but also manage network functions autonomously, both before and after deployment. In this paper, we propose a comprehensive framework for reasoning-and-acting (ReAct) agents powered by multimodal large language models (LLMs), facilitating seamless translation of user intent into network actions. The proposed architecture integrates multimodal retrieval-augmented generation (RAG), persona-driven reasoning, and function calling to orchestrate tasks across vision, text, and radio modalities. We examine how RAG influences reasoning in multimodal agents, how LLMs can be specialised for telecom applications, and which performance benchmarks should guide their design. A UAV-based coverage optimisation case study demonstrates the framework’s ability to leverage vision data to enhance average signal-to-noise ratio (SINR), yielding substantial reductions in outage probability and improvements in throughput under Rician fading channels. Analytical characterisation and practical implementation results are presented to benchmark performance.

Förderung

  • Advanced Connectivity Technology Hub, Communications Hub for Empowering Distributed Cloud Computing Applications and Research (CHEDDAR) through the Engineering…
  • Academic Grant from Google Inc. for exploiting the GenAI capabilities in VertexAI
  • Department for Science, Innovation and Technology (DSIT), the U.K. AI Security Institute (AISI), and EPSRC
  • UKRI Funding Service through Strategic Decision-Making and Cooperation among AI Agents: Exploring Safety and Governance in Telecom

Zitieren

APA

Zaidi, S. A. R., Hafeez, M., Qazzaz, M. M. H., Tatipamula, M., Chowdhury, K., & Win, M. Z. (2026). Reasoning and Acting (ReAct) With Multimodal LLMs: A Framework for Intent Driven 6G Networks. IEEE Open Journal of the Communications Society, 7, 5463-5481. https://doi.org/10.1109/ojcoms.2026.3693067

Vancouver

Zaidi SAR, Hafeez M, Qazzaz MMH, Tatipamula M, Chowdhury K, Win MZ. Reasoning and Acting (ReAct) With Multimodal LLMs: A Framework for Intent Driven 6G Networks. IEEE Open Journal of the Communications Society. 2026;7:5463-5481. doi:10.1109/ojcoms.2026.3693067

BibTeX
@article{Zaidi2026,
  author = {Zaidi, Syed Ali Raza and Hafeez, Maryam and Qazzaz, Mohammed M. H. and Tatipamula, Mallik and Chowdhury, Kuntal and Win, Moe Z.},
  title = {Reasoning and Acting (ReAct) With Multimodal LLMs: A Framework for Intent Driven 6G Networks},
  journal = {IEEE Open Journal of the Communications Society},
  year = {2026},
  volume = {7},
  pages = {5463--5481},
  doi = {10.1109/ojcoms.2026.3693067}
}
Kostenlos starten →