Self-Adaptive Multi-Agent LLM-Based Security Pattern Selection for IoT Systems

Saeid Jamshidi, Foutse Khomh, Carol Fung, Kawser Wazed Nafi · 2026

Read Self-Adaptive Multi-Agent LLM-Based Security Pattern Selection for IoT Systems by Saeid Jamshidi, Foutse Khomh, Carol Fung, Kawser Wazed Nafi (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-2605-00741.txt.

Description

Saeid Jamshidi and 3 co-authors let language-model agents choose security controls for Internet of Things systems, while a deterministic core keeps them safe. Edge IoT deployments need defenses that adapt at runtime, but choosing and combining mitigation actions under latency, energy, and computing limits, without unsafe interactions between controls, is a constrained decision problem. Static rule sets and learned policies offer few guarantees about feasibility or conflict safety.

Their system, ASPO, separates proposal from enforcement inside a standard self-adaptation loop. Language-model agents suggest portfolios of security patterns, and a deterministic optimizer then enforces action integrity, conflict-free composition, and resource feasibility at every decision step. Deployed on an edge-gateway testbed against replayed attack traffic, the system activated controls with perfect conflict freedom and consistent feasibility. Deeper exploration compressed the worst-case latency and energy overheads by about a fifth, at no increase in average energy use.

Catalog record

CREATORSaeid Jamshidi, Foutse Khomh, Carol Fung, Kawser Wazed Nafi
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE77 KB
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2605.00741

Browse the full library · About the Tower of Babel