Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

LLM-Based Net Analyzer rApp for Explainable and Safe Automation in O-RAN Non-RT RIC

Tuan V. Ngo, Mao V. Ngo, Binbin Chen, Tony Q. S. Quek, Tejaswita Kumari, Maziar Nekovee

Abstract

Modern 5G/6G radio access networks are increasingly programmable through O-RAN, yet their operational complexity has grown with disaggregation, open interfaces, and fine-grained control parameters. While RAN-side analytics and telemetry mechanisms, such as KPI-based monitoring and mobility event reporting, provide visibility into network behavior, operators still face challenges in correlating heterogeneous events and safely translating observations into actionable configuration changes. This paper presents an LLM-based Net Analyzer rApp for the O-RAN Non-RT RIC that enables explainable and safe, human-in-the-loop automation for RAN operations. The proposed rApp adopts an event-informed, batch-triggered reasoning framework in which mobility events are first interpreted, anomalies are confirmed through targeted log inspection, configurations are inspected via tool-gated access, and minimal configuration changes are proposed only after explicit operator approval. The architecture enforces a strict separation between reasoning and actuation, ensuring auditability and operational safety. The system is implemented and demonstrated on a real O-RAN testbed using a reproducible ping-pong handover scenario, illustrating how large language models can function as reasoning co-pilots that transform raw RAN telemetry into structured explanations and controlled remediation workflows, complementing existing analytics-only approaches in the NonRT RIC.

LLM-Based Net Analyzer rApp for Explainable and Safe Automation in O-RAN Non-RT RIC

Abstract

Modern 5G/6G radio access networks are increasingly programmable through O-RAN, yet their operational complexity has grown with disaggregation, open interfaces, and fine-grained control parameters. While RAN-side analytics and telemetry mechanisms, such as KPI-based monitoring and mobility event reporting, provide visibility into network behavior, operators still face challenges in correlating heterogeneous events and safely translating observations into actionable configuration changes. This paper presents an LLM-based Net Analyzer rApp for the O-RAN Non-RT RIC that enables explainable and safe, human-in-the-loop automation for RAN operations. The proposed rApp adopts an event-informed, batch-triggered reasoning framework in which mobility events are first interpreted, anomalies are confirmed through targeted log inspection, configurations are inspected via tool-gated access, and minimal configuration changes are proposed only after explicit operator approval. The architecture enforces a strict separation between reasoning and actuation, ensuring auditability and operational safety. The system is implemented and demonstrated on a real O-RAN testbed using a reproducible ping-pong handover scenario, illustrating how large language models can function as reasoning co-pilots that transform raw RAN telemetry into structured explanations and controlled remediation workflows, complementing existing analytics-only approaches in the NonRT RIC.
Paper Structure (12 sections, 6 figures)

This paper contains 12 sections, 6 figures.

Table of Contents

  1. Introduction
  2. Related Work
  3. LLM-Based Net Analyzer rApp Architecture and Design
  4. System Architecture
  5. LLM-Based Reasoning Framework
  6. Event-Informed, Batch-Triggered Automation Loop
  7. Experimental Setup and Evaluation
  8. O-RAN Testbed Description
  9. Ping-Pong Handover Experiment Setup
  10. End-to-End rApp Execution
  11. Experimental Results and Discussion
  12. Conclusion

Figures (6)

  • Figure 1: Simplified architecture of the LLM-based Net Analyzer rApp. RAN event inputs and human chat inputs are processed by an LLM-based reasoning agent operating within an explicit event-informed, batch-triggered loop.
  • Figure 2: O-RAN testbed and network topology used for evaluating the proposed Net Analyzer rApp. Two gNBs (gNB-30, gNB-31), each connected to a corresponding RU, form two neighboring indoor cells with partial coverage overlap. The Non-RT RIC hosts multiple components, including events, log and configuration collectors, data storage backends, and the proposed rApp.
  • Figure 3: Ping-pong handover experimental scenario: A 5G robot-dog follows the indicated trajectory, repeatedly crossing the cell boundary and triggering successive handovers between the two cells.
  • Figure 4: Simplified dialogue between Net Analyzer rApp and Human Operator during resolving the ping-pong handover issue.
  • Figure 5: Video streaming frame per second (FPS) comparison between the two scenarios with/without the proposed rApp enabled.
  • ...and 1 more figures