Multi-agent LLMs on the Cloud-Edge-IoT Continuum: A Systematic Mapping Study on Architectures, Deployment, and Evaluation
- Published
- Source
- arXiv
- Paper number
- 1359
- Field
- Distributed Systems
- arXiv ID
- 2610.28702
Key points
- From 831 records, the authors selected 45 primary studies and classified them by architecture family, deployment pattern, application domain, and evaluation practice.
- Five architecture families recur in the corpus, and deployment spans 6 patterns, ranging from local-only execution to control loops gated by policy checks or human approval.
- 24 of the 45 studies address IoT, industrial, smart-home, healthcare, vehicular, or robotics settings, 29 evaluate on physical hardware, and open models below 8B parameters dominate that hardware.
- Only 6 studies release code or data, and only 15 of the 45 report ablation experiments.
- Privacy motivates many local designs, but privacy is directly measured in only 2 studies.
- The authors provide a reporting checklist asking studies to report task, continuum, state, and governance conditions together with outcomes for comparability.
Paper links
External research summaries. These are not HDATF publications or measured product results.