Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

End-to-End Edge AI Service Provisioning Framework in 6G ORAN

Yun Tang, Udhaya Chandhar Srinivasan, Benjamin James Scott, Obumneme Umealor, Dennis Kevogo, Weisi Guo

TL;DR

The paper tackles end-to-end provisioning of edge AI services in 6G O-RAN by introducing an LLM-driven orchestrator that maps high-level intents to AI service deployment, model selection, network adaptation, and QoS monitoring. It integrates an AI model registry (e.g., HuggingFace), edge deployment on MEC, and cross-domain network control through SBI/PCF, with monitoring via xApps to predict and prevent QoS degradations. A prototype based on OpenAirInterface, FlexRIC, LangGraph, Gemini, and HuggingFace demonstrates the seamless end-to-end workflow from user intent to deployment and adaptive network behavior, completed in minutes on a single testbed. The work highlights the potential of AI-native orchestration to unify service and network management in future 6G architectures, enabling scalable and autonomous edge intelligence, while acknowledging avenues for real-time validation, scalability, and multi-agent enhancements.

Abstract

With the advent of 6G, Open Radio Access Network (O-RAN) architectures are evolving to support intelligent, adaptive, and automated network orchestration. This paper proposes a novel Edge AI and Network Service Orchestration framework that leverages Large Language Model (LLM) agents deployed as O-RAN rApps. The proposed LLM-agent-powered system enables interactive and intuitive orchestration by translating the user's use case description into deployable AI services and corresponding network configurations. The LLM agent automates multiple tasks, including AI model selection from repositories (e.g., Hugging Face), service deployment, network adaptation, and real-time monitoring via xApps. We implement a prototype using open-source O-RAN projects (OpenAirInterface and FlexRIC) to demonstrate the feasibility and functionality of our framework. Our demonstration showcases the end-to-end flow of AI service orchestration, from user interaction to network adaptation, ensuring Quality of Service (QoS) compliance. This work highlights the potential of integrating LLM-driven automation into 6G O-RAN ecosystems, paving the way for more accessible and efficient edge AI ecosystems.

End-to-End Edge AI Service Provisioning Framework in 6G ORAN

TL;DR

The paper tackles end-to-end provisioning of edge AI services in 6G O-RAN by introducing an LLM-driven orchestrator that maps high-level intents to AI service deployment, model selection, network adaptation, and QoS monitoring. It integrates an AI model registry (e.g., HuggingFace), edge deployment on MEC, and cross-domain network control through SBI/PCF, with monitoring via xApps to predict and prevent QoS degradations. A prototype based on OpenAirInterface, FlexRIC, LangGraph, Gemini, and HuggingFace demonstrates the seamless end-to-end workflow from user intent to deployment and adaptive network behavior, completed in minutes on a single testbed. The work highlights the potential of AI-native orchestration to unify service and network management in future 6G architectures, enabling scalable and autonomous edge intelligence, while acknowledging avenues for real-time validation, scalability, and multi-agent enhancements.

Abstract

With the advent of 6G, Open Radio Access Network (O-RAN) architectures are evolving to support intelligent, adaptive, and automated network orchestration. This paper proposes a novel Edge AI and Network Service Orchestration framework that leverages Large Language Model (LLM) agents deployed as O-RAN rApps. The proposed LLM-agent-powered system enables interactive and intuitive orchestration by translating the user's use case description into deployable AI services and corresponding network configurations. The LLM agent automates multiple tasks, including AI model selection from repositories (e.g., Hugging Face), service deployment, network adaptation, and real-time monitoring via xApps. We implement a prototype using open-source O-RAN projects (OpenAirInterface and FlexRIC) to demonstrate the feasibility and functionality of our framework. Our demonstration showcases the end-to-end flow of AI service orchestration, from user interaction to network adaptation, ensuring Quality of Service (QoS) compliance. This work highlights the potential of integrating LLM-driven automation into 6G O-RAN ecosystems, paving the way for more accessible and efficient edge AI ecosystems.

Paper Structure

This paper contains 15 sections, 3 figures.

Table of Contents

  1. Introduction
  2. Related Works
  3. Edge AI Orchestration Approaches
  4. O-RAN and 6G
  5. LLMs in 6G Networks
  6. The Framework
  7. AI Model Discovery and Selection
  8. AI Service Deployment
  9. Network Adaptation Mechanism
  10. Service Monitoring and Prediction
  11. Demonstration
  12. Prototype Implementation
  13. End-to-End AI Service Provisioning Workflow
  14. Discussion
  15. Conclusion

Figures (3)

  • Figure 1: Edge AI use cases in 6G Network-powered modern society.
  • Figure 2: End-to-End Edge AI Service Provisioning Framework Overview.
  • Figure 3: Agent tasks workflow for the demonstration use case.