Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Large Language Models meet Network Slicing Management and Orchestration

Abdulhalim Dandoush, Viswanath Kumarskandpriya, Mueen Uddin, Usman Khalil

TL;DR

The paper addresses the challenge of managing end-to-end network slices across multiple administrative domains by proposing a vision in which Large Language Models (LLMs) and multi-agent systems translate user intent into technical requirements, map network functions to infrastructure, and automate the slice lifecycle within existing MANO frameworks. It introduces an LLM-based, multi-agent framework that aligns with ETSI/3GPP management functions, enabling collaboration across domains and abstraction of topology to support cross-domain slice deployment and monitoring. Key contributions include a structured agent architecture, user- and operator-facing workflow outlines, and a discussion of challenges (data, security, trust, interoperability) along with forward-looking solutions (federated learning, explainability, continuous auditing). The proposed approach aims to enhance automation, agility, and resource efficiency in multi-domain networks, with potential impact on operators, tenants, and service innovation in future 5G/6G ecosystems. All mathematical notation, where present, is intended to be clearly delimited.

Abstract

Network slicing, a cornerstone technology for future networks, enables the creation of customized virtual networks on a shared physical infrastructure. This fosters innovation and agility by providing dedicated resources tailored to specific applications. However, current orchestration and management approaches face limitations in handling the complexity of new service demands within multi-administrative domain environments. This paper proposes a future vision for network slicing powered by Large Language Models (LLMs) and multi-agent systems, offering a framework that can be integrated with existing Management and Orchestration (MANO) frameworks. This framework leverages LLMs to translate user intent into technical requirements, map network functions to infrastructure, and manage the entire slice lifecycle, while multi-agent systems facilitate collaboration across different administrative domains. We also discuss the challenges associated with implementing this framework and potential solutions to mitigate them.

Large Language Models meet Network Slicing Management and Orchestration

TL;DR

The paper addresses the challenge of managing end-to-end network slices across multiple administrative domains by proposing a vision in which Large Language Models (LLMs) and multi-agent systems translate user intent into technical requirements, map network functions to infrastructure, and automate the slice lifecycle within existing MANO frameworks. It introduces an LLM-based, multi-agent framework that aligns with ETSI/3GPP management functions, enabling collaboration across domains and abstraction of topology to support cross-domain slice deployment and monitoring. Key contributions include a structured agent architecture, user- and operator-facing workflow outlines, and a discussion of challenges (data, security, trust, interoperability) along with forward-looking solutions (federated learning, explainability, continuous auditing). The proposed approach aims to enhance automation, agility, and resource efficiency in multi-domain networks, with potential impact on operators, tenants, and service innovation in future 5G/6G ecosystems. All mathematical notation, where present, is intended to be clearly delimited.

Abstract

Network slicing, a cornerstone technology for future networks, enables the creation of customized virtual networks on a shared physical infrastructure. This fosters innovation and agility by providing dedicated resources tailored to specific applications. However, current orchestration and management approaches face limitations in handling the complexity of new service demands within multi-administrative domain environments. This paper proposes a future vision for network slicing powered by Large Language Models (LLMs) and multi-agent systems, offering a framework that can be integrated with existing Management and Orchestration (MANO) frameworks. This framework leverages LLMs to translate user intent into technical requirements, map network functions to infrastructure, and manage the entire slice lifecycle, while multi-agent systems facilitate collaboration across different administrative domains. We also discuss the challenges associated with implementing this framework and potential solutions to mitigate them.
Paper Structure (13 sections, 7 figures)

This paper contains 13 sections, 7 figures.

Table of Contents

  1. Introduction
  2. Large Language Models Overview
  3. Basic E2E Network Slicing Architecture
  4. Limitations and Challenges in Applying Network Slicing Framework
  5. Proposed LLM-based Slicing Management approach
  6. From Customer request to auto deployment
  7. Service Consumer point of view
  8. Network Operator point of view
  9. Challenges and Future Directions
  10. Current Challenges
  11. Future Directions and Solutions
  12. Conclusion
  13. Sample LLM Agents Definition with agent role, description and associated tools used by each agent.

Figures (7)

  • Figure 1: SDO's views on Network Slicing - GSMA, 3GPP and ETSI
  • Figure 2: LLM Agents scope mapped to 3GPP and ETSI related management functions responsible for network slice deployments
  • Figure 3: High level view of LLMs Assisted Network Slice Management
  • Figure 4: LLM Assisted End User Network Slice Intent Translation
  • Figure 5: A hypothetical interaction between an end user and an intelligent LLM, negotiating over a telemedicine service
  • ...and 2 more figures