RAN-Level Open Questions on ISAC Life-Cycle, Real-Time Data Access, and Multi-Node Coordination

Determine RAN-level mechanisms within the O-RAN/3GPP architecture to support (i) dynamic life-cycle management of integrated sensing and communication (ISAC) and AI inference algorithms, (ii) real-time exposure of user-plane I/Q, CSI, and SRS data at sub-millisecond sensing timescales, and (iii) multi-node sensing coordination across base stations.

Background

The paper surveys prior architectural work on integrating sensing into cellular networks and observes that much of it concentrates on single-node topologies, waveform design, or core-network interfaces. The authors argue that this leaves key system-level issues unresolved at the RAN level—specifically algorithm life-cycle management, real-time user-plane data access for sensing, and multi-node sensing coordination.

They propose dApps at the DU and an E3 interface to expose I/Q-level data, combined with hierarchical control via xApps/rApps, as a path forward. However, they explicitly flag that these aspects are still open questions to be addressed by the community.

References

Prior architectural work has explored sensing integration in cellular networks, but has largely focused on single-node topologies, waveform design, or core-network interfaces—leaving open questions on algorithm life-cycle management, real-time I/Q-level data access, and multi-node sensing coordination at the RAN level.

— Enabling Programmable Inference and ISAC at the 6GR Edge with dApps  (2603.29146 - Polese et al., 31 Mar 2026) in Section 1, Introduction