Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Towards an AI/ML-driven SMO Framework in O-RAN: Scenarios, Solutions, and Challenges

Mohammad Asif Habibi, Bin Han, Merve Saimler, Ignacio Labrador Pavon, Hans D. Schotten

TL;DR

This manuscript proposes three scenarios for integrating machine learning (ML) algorithms into SMO and focuses on exploring one of the scenarios in which the non-real-time RAN intelligence controller (Non-RT RIC) plays a major role in data collection, as well as model training, deployment, and refinement, by proposing a centralized ML architecture.

Abstract

The emergence of the open radio access network (O-RAN) architecture offers a paradigm shift in cellular network management and service orchestration, leveraging data-driven, intent-based, autonomous, and intelligent solutions. Within O-RAN, the service management and orchestration (SMO) framework plays a pivotal role in managing network functions (NFs), resource allocation, service provisioning, and others. However, the increasing complexity and scale of O-RANs demand autonomous and intelligent models for optimizing SMO operations. To achieve this goal, it is essential to integrate intelligence and automation into the operations of SMO. In this manuscript, we propose three scenarios for integrating machine learning (ML) algorithms into SMO. We then focus on exploring one of the scenarios in which the non-real-time RAN intelligence controller (Non-RT RIC) plays a major role in data collection, as well as model training, deployment, and refinement, by proposing a centralized ML architecture. Finally, we identify potential challenges associated with implementing a centralized ML solution within SMO.

Towards an AI/ML-driven SMO Framework in O-RAN: Scenarios, Solutions, and Challenges

TL;DR

This manuscript proposes three scenarios for integrating machine learning (ML) algorithms into SMO and focuses on exploring one of the scenarios in which the non-real-time RAN intelligence controller (Non-RT RIC) plays a major role in data collection, as well as model training, deployment, and refinement, by proposing a centralized ML architecture.

Abstract

The emergence of the open radio access network (O-RAN) architecture offers a paradigm shift in cellular network management and service orchestration, leveraging data-driven, intent-based, autonomous, and intelligent solutions. Within O-RAN, the service management and orchestration (SMO) framework plays a pivotal role in managing network functions (NFs), resource allocation, service provisioning, and others. However, the increasing complexity and scale of O-RANs demand autonomous and intelligent models for optimizing SMO operations. To achieve this goal, it is essential to integrate intelligence and automation into the operations of SMO. In this manuscript, we propose three scenarios for integrating machine learning (ML) algorithms into SMO. We then focus on exploring one of the scenarios in which the non-real-time RAN intelligence controller (Non-RT RIC) plays a major role in data collection, as well as model training, deployment, and refinement, by proposing a centralized ML architecture. Finally, we identify potential challenges associated with implementing a centralized ML solution within SMO.
Paper Structure (17 sections, 2 figures)

This paper contains 17 sections, 2 figures.

Table of Contents

  1. Introductory Remarks
  2. Deployment Scenarios for Integrating MDA into the SMO Framework
  3. Scenario A: External Data or Model Importation
  4. Scenario B: Internal Model Training and Deployment
  5. Scenario C: Collaborative Data or Model Sharing
  6. Centralized ML Model Training and Deployment Solutions within SMO
  7. Data Collection and Preprocessing
  8. Model Training and Development
  9. Model Deployment and Inference
  10. Major Research Challenges
  11. Privacy
  12. Resilience
  13. Elasticity
  14. Agility
  15. Signaling Efficiency
  16. ...and 2 more sections

Figures (2)

  • Figure 1: Three scenarios for incorporating AI/ML models into SMO are depicted. The scopes of Scenario A, Scenario B, and Scenario C are portrayed within the dashed blue, red, and green boxes, respectively. The legends contained within this figure have been defined in our previous work BuildingSMOFramework.
  • Figure 2: Proposed workflow for centralized ML model training, deployment, and refinement within the SMO framework of O-RAN architecture