α 3 \alpha^{3} -Bench: A Unified Benchmark of Safety, Robustness, and Efficiency for LLM-Based UAV Agents over 6G Networks
Mohamed Amine Ferrag, Abderrahmane Lakas, and Merouane Debbah · 2026
Read α 3 \alpha^{3} -Bench: A Unified Benchmark of Safety, Robustness, and Efficiency for LLM-Based UAV Agents over 6G Networks by Mohamed Amine Ferrag, Abderrahmane Lakas, and Merouane Debbah (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2601-03281.txt.
Description
Mohamed Amine Ferrag, Abderrahmane Lakas, and Merouane Debbah introduce a benchmark for language-model agents that fly drones over future 6G networks. Such agents increasingly act as high-level controllers for unmanned aerial missions, yet evaluations rarely test whether they stay safe, protocol compliant, and effective under realistic network strain. Their benchmark frames each mission as a multi-turn conversation between the drone agent and a human operator.
Inside that loop, decisions must respect strict message schemas, mission policies, speaker alternation, and safety constraints. All this happens while the network shifts beneath them: slices change, latency and jitter swing, packets drop, and edge load varies. A dual action layer supports both tool calls and agent-to-agent coordination. The benchmark is built on a large corpus of about one hundred thirteen thousand conversational cases.
Catalog record
| CREATOR | Mohamed Amine Ferrag, Abderrahmane Lakas, and Merouane Debbah |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 102 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2601-03281