Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

5G-CT: Automated Deployment and Over-the-Air Testing of End-to-End Open Radio Access Networks

Leonardo Bonati, Michele Polese, Salvatore D'Oro, Pietro Brach del Prever, Tommaso Melodia

TL;DR

The paper addresses the complexity of deploying and testing end-to-end Open RAN/5G systems, where manual interventions slow innovation. It presents 5G-CT, a cloud-native framework built on Red Hat OpenShift that utilizes GitOps, Tekton, and ArgoCD to automatically deploy and test disaggregated 5G/Open RAN stacks, including OpenAirInterface and commercial cores, via over-the-air experiments. Key contributions include designing automated pipelines for test specification, repeatable builds, automatic open-RAN testing, and supporting services; implementing radio/spectrum management microservices; and validating the approach with months of OTA testing on an Arena testbed. The results demonstrate scalable, low-latency deployment and continuous validation of multi-vendor components, enabling more reliable, robust, and secure Open RAN software with reduced human effort and faster iteration cycles.

Abstract

Deploying and testing cellular networks is a complex task due to the multitude of components involved -- from the core to the Radio Access Network (RAN) and User Equipment (UE) -- all of which requires integration and constant monitoring. Additional challenges are posed by the nature of the wireless channel, whose inherent randomness hinders the repeatability and consistency of the testing process. Consequently, existing solutions for both private and public cellular systems still rely heavily on human intervention for operations such as network reconfiguration, performance monitoring, and end-to-end testing. This reliance significantly slows the pace of innovation in cellular systems. To address these challenges, we introduce 5G-CT, an automation framework based on OpenShift and the GitOps workflow, capable of deploying a softwarized end-to-end 5G and O-RAN-compliant system in a matter of seconds without the need for any human intervention. We have deployed 5G-CT to test the integration and performance of open-source cellular stacks, including OpenAirInterface, and have collected months of automated over-the-air testing results involving software-defined radios. 5G-CT brings cloud-native continuous integration and delivery to the RAN, effectively addressing the complexities associated with managing spectrum, radios, heterogeneous devices, and distributed components. Moreover, it endows cellular networks with much needed automation and continuous testing capabilities, providing a platform to evaluate the robustness and resiliency of Open RAN software.

5G-CT: Automated Deployment and Over-the-Air Testing of End-to-End Open Radio Access Networks

TL;DR

The paper addresses the complexity of deploying and testing end-to-end Open RAN/5G systems, where manual interventions slow innovation. It presents 5G-CT, a cloud-native framework built on Red Hat OpenShift that utilizes GitOps, Tekton, and ArgoCD to automatically deploy and test disaggregated 5G/Open RAN stacks, including OpenAirInterface and commercial cores, via over-the-air experiments. Key contributions include designing automated pipelines for test specification, repeatable builds, automatic open-RAN testing, and supporting services; implementing radio/spectrum management microservices; and validating the approach with months of OTA testing on an Arena testbed. The results demonstrate scalable, low-latency deployment and continuous validation of multi-vendor components, enabling more reliable, robust, and secure Open RAN software with reduced human effort and faster iteration cycles.

Abstract

Deploying and testing cellular networks is a complex task due to the multitude of components involved -- from the core to the Radio Access Network (RAN) and User Equipment (UE) -- all of which requires integration and constant monitoring. Additional challenges are posed by the nature of the wireless channel, whose inherent randomness hinders the repeatability and consistency of the testing process. Consequently, existing solutions for both private and public cellular systems still rely heavily on human intervention for operations such as network reconfiguration, performance monitoring, and end-to-end testing. This reliance significantly slows the pace of innovation in cellular systems. To address these challenges, we introduce 5G-CT, an automation framework based on OpenShift and the GitOps workflow, capable of deploying a softwarized end-to-end 5G and O-RAN-compliant system in a matter of seconds without the need for any human intervention. We have deployed 5G-CT to test the integration and performance of open-source cellular stacks, including OpenAirInterface, and have collected months of automated over-the-air testing results involving software-defined radios. 5G-CT brings cloud-native continuous integration and delivery to the RAN, effectively addressing the complexities associated with managing spectrum, radios, heterogeneous devices, and distributed components. Moreover, it endows cellular networks with much needed automation and continuous testing capabilities, providing a platform to evaluate the robustness and resiliency of Open RAN software.
Paper Structure (7 sections, 5 figures)

This paper contains 7 sections, 5 figures.

Table of Contents

  1. Introduction
  2. Related Work
  3. Automated Open RAN Pipelines
  4. CI, CD, and CT of Open RAN gNBs
  5. 5G-CT OpenShift-based Automated Infrastructure and Microservices
  6. Experimental Results
  7. Conclusions

Figures (5)

  • Figure 1: Intent-based CI/CD/CT pipeline for Open RAN.
  • Figure 2: Example of pipeline for CI/CD/CT of Open RAN gNB.
  • Figure 3: 5G-CT architecture.
  • Figure 4: Time evolution of throughput (box plots) and packet loss (bar plot) of an OAI gNB during automated tests performed over 9 months for different data rates. The vertical line in October 2023 marks the switch from real-time to generic kernel of the host machines running the gNB.
  • Figure 5: Comparison between unsuccessful and successful tests of an OAI gNB for different targeted data rates. The solid lines show the history of previous tests, with shaded areas depicting their confidence intervals. The dashed lines show the most recent test to compare.