Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

ML-based handover prediction over a real O-RAN deployment using RAN Intelligent controller

Merim Dzaferagic, Bruno Missi Xavier, Diarmuid Collins, Vince D'Onofrio, Magnos Martinello, Marco Ruffini

TL;DR

This work tackles predicting handover events in a real O-RAN deployment to enable dynamic resource acquisition. It adopts an LSTM-based sequence model trained on OpenIreland data to forecast handovers using standard RAN measurements, and links model performance to operational costs. The study demonstrates recall up to 88% with a minimum precision of 75% and reports over 80% cost savings versus traditional long-term resource purchases, highlighting the practical impact for multiple network stakeholders. It also outlines future directions including federated and transfer learning to improve privacy and adaptability across environments.

Abstract

O-RAN introduces intelligent and flexible network control in all parts of the network. The use of controllers with open interfaces allow us to gather real time network measurements and make intelligent/informed decision. The work in this paper focuses on developing a use-case for open and reconfigurable networks to investigate the possibility to predict handover events and understand the value of such predictions for all stakeholders that rely on the communication network to conduct their business. We propose a Long-Short Term Memory Machine Learning approach that takes standard Radio Access Network measurements to predict handover events. The models were trained on real network data collected from a commercial O-RAN setup deployed in our OpenIreland testbed. Our results show that the proposed approach can be optimized for either recall or precision, depending on the defined application level objective. We also link the performance of the Machine Learning (ML) algorithm to the network operation cost. Our results show that ML-based matching between the required and available resources can reduce operational cost by more than 80%, compared to long term resource purchases.

ML-based handover prediction over a real O-RAN deployment using RAN Intelligent controller

TL;DR

This work tackles predicting handover events in a real O-RAN deployment to enable dynamic resource acquisition. It adopts an LSTM-based sequence model trained on OpenIreland data to forecast handovers using standard RAN measurements, and links model performance to operational costs. The study demonstrates recall up to 88% with a minimum precision of 75% and reports over 80% cost savings versus traditional long-term resource purchases, highlighting the practical impact for multiple network stakeholders. It also outlines future directions including federated and transfer learning to improve privacy and adaptability across environments.

Abstract

O-RAN introduces intelligent and flexible network control in all parts of the network. The use of controllers with open interfaces allow us to gather real time network measurements and make intelligent/informed decision. The work in this paper focuses on developing a use-case for open and reconfigurable networks to investigate the possibility to predict handover events and understand the value of such predictions for all stakeholders that rely on the communication network to conduct their business. We propose a Long-Short Term Memory Machine Learning approach that takes standard Radio Access Network measurements to predict handover events. The models were trained on real network data collected from a commercial O-RAN setup deployed in our OpenIreland testbed. Our results show that the proposed approach can be optimized for either recall or precision, depending on the defined application level objective. We also link the performance of the Machine Learning (ML) algorithm to the network operation cost. Our results show that ML-based matching between the required and available resources can reduce operational cost by more than 80%, compared to long term resource purchases.
Paper Structure (11 sections, 11 equations, 9 figures, 1 table)

This paper contains 11 sections, 11 equations, 9 figures, 1 table.

Table of Contents

  1. Introduction
  2. Problem Description and Experimental Setup
  3. Problem Description
  4. Testbed facilities
  5. Experimental setup
  6. xApp Design
  7. Data Collection
  8. ML workflow
  9. Stakeholder adaptation
  10. ML Model Performance Evaluation
  11. Conclusions

Figures (9)

  • Figure 1: O-RAN architecture showing that the functional parts of the CU/DU can be moved closer to and further from the RU.
  • Figure 2: Overall architecture of the OpenIreland testbed.
  • Figure 3: Resource allocation areas, located around the Trinity College Dublin campus.
  • Figure 4: The benefits various stakeholders gain from the use of information provided by an O-RAN intelligent control mechanism (i.e. xApp).
  • Figure 5: Recall, precision and F1-score depending on the prediction horizon. The prediction history is fixed at $10s$.
  • ...and 4 more figures