wireless networks

AIMS: An Agentic AI Framework for Sim-to-Real Multi-Modal ISAC

Authors

Yijie Bian, Kai Zhang, Wei Guo, Zixin Wang, Shenghui Song, Jun Zhang, Khaled B. Letaief

Venue

arXiv

Published

2026-09-30

ID

arXiv:2609.39964

Proposes AIMS, an agentic AI framework that turns a plain-language deployment request into a sim-to-real configuration for multi-modal integrated sensing and communication (ISAC) in wireless networks. A scene construction agent and a scene understanding agent coordinate sensing, wireless and learning components, and experiments on the real-world DeepSense 6G dataset show improved vehicle detection and beam prediction over simulation and fusion baselines.