Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

A Vision-based Framework for Intelligent gNodeB Mobility Control

Pedro Duarte, André Coelho, Francisco Ribeiro, Filipe B. Teixeira, Luís Pessoa, Manuel Ricardo

Abstract

This paper proposes a vision-based framework for the intelligent control of mobile Open Radio Access Network (O-RAN) base stations (gNBs) operating in dynamic wireless environments. The framework comprises three innovative components. The first is the introduction of novel Service Models (SMs) within a vision-enabled O-RAN architecture, termed VisionRAN. These SMs extend state-of-the-art O-RAN-based architectures by enabling the transmission of vision-based sensing data and gNB positioning control messages. The second is an O-RAN xApp, VisionApp, which fuses vision and radio data, and uses this information to control the position of a mobile gNB, using a Deep Q-Network (DQN). The third is a digital twin environment, VisionTwin, which incorporates vision data and can emulate realistic wireless scenarios; this digital twin was used to train the DQN running in VisionApp and validate the overall system. Experimental results, obtained using real vision data and an emulated radio, demonstrate that the proposed approach reduces the duration of Line-of-Sight (LoS) blockages by up to 75% compared to a static gNB. These findings confirm the viability of integrating multimodal perception and learning-based control within RANs.

A Vision-based Framework for Intelligent gNodeB Mobility Control

Abstract

This paper proposes a vision-based framework for the intelligent control of mobile Open Radio Access Network (O-RAN) base stations (gNBs) operating in dynamic wireless environments. The framework comprises three innovative components. The first is the introduction of novel Service Models (SMs) within a vision-enabled O-RAN architecture, termed VisionRAN. These SMs extend state-of-the-art O-RAN-based architectures by enabling the transmission of vision-based sensing data and gNB positioning control messages. The second is an O-RAN xApp, VisionApp, which fuses vision and radio data, and uses this information to control the position of a mobile gNB, using a Deep Q-Network (DQN). The third is a digital twin environment, VisionTwin, which incorporates vision data and can emulate realistic wireless scenarios; this digital twin was used to train the DQN running in VisionApp and validate the overall system. Experimental results, obtained using real vision data and an emulated radio, demonstrate that the proposed approach reduces the duration of Line-of-Sight (LoS) blockages by up to 75% compared to a static gNB. These findings confirm the viability of integrating multimodal perception and learning-based control within RANs.
Paper Structure (16 sections, 1 equation, 7 figures, 4 tables)

This paper contains 16 sections, 1 equation, 7 figures, 4 tables.

Table of Contents

  1. Introduction
  2. Vision-Enabled O-RAN Architecture
  3. New Vision Service Models for Multimodal Perception
  4. E2 Agents for Multimodal Integration
  5. VisionRAN Architecture
  6. Vision-Aided xApp for gNB Mobility Control
  7. Vision-Oriented Digital Twin Environment
  8. Training of Intelligent Models for Vision-Based gNB Control
  9. Emulation of gNB Perception and Mobility Control
  10. Experimental Evaluation
  11. Physical scenario
  12. Positioning accuracy evaluation
  13. Network performance evaluation
  14. Discussion
  15. Related Work
  16. ...and 1 more sections

Figures (7)

  • Figure 1: Proposed VisionRAN architecture.
  • Figure 2: Deep Q-Network architecture.
  • Figure 3: Overview of the algorithmic logic of VisionApp within near-RT RIC. VisionApp operates with a control interval of 200 ms, which is aligned with the timing constraints of near-RT RIC applications. All processing stages are lightweight and designed to execute within this interval, ensuring practical deployability in real-time O-RAN control loops.
  • Figure 4: Emulation of gNB perception and mobility control using VisionTwin.
  • Figure 5: Physical scenario, including an RGB-D video camera and a static obstacle, with annotated reference positions used as ground truth.
  • ...and 2 more figures