PRAXEUM

PRAXEUMPRAXEUMPRAXEUM

PRAXEUM

PRAXEUMPRAXEUMPRAXEUM
  • Home
  • Curriculum
    • Course Description
    • Coursework
    • Class Slides
    • Assessments
    • YANG Models
    • Additional Resources
  • More
    • Home
    • Curriculum
      • Course Description
      • Coursework
      • Class Slides
      • Assessments
      • YANG Models
      • Additional Resources
  • Home
  • Curriculum
    • Course Description
    • Coursework
    • Class Slides
    • Assessments
    • YANG Models
    • Additional Resources

In Week 2, we introduced the SMO and O-Cloud briefly. This week, we’ll go deeper into how virtualization and cloud-native principles make those systems more powerful, scalable, and efficient.

Week 4: Virtualization and Cloudification in Open RAN

Introduction to Network Function Virtualization (NFV) in Open RAN

What is Network Function Virtualization (NFV)?

  • NFV means running network tasks (like routing or signal processing) as software instead of using special hardware.
  • DU and CU functions can run on regular servers in a data center.

Layers of NFV Architecture

  1. Hardware: General-purpose servers (compute, storage, networking)
  2. Virtualization Layer: Software (like hypervisors or containers) that runs network functions
  3. VNFs or CNFs: Virtual or cloud-native functions (like the CU)
  4. MANO: Management and orchestration layer that controls everything

Cloud RAN vs. Cloud-Native RAN

  • C-RAN: Moves some RAN functions into a centralized cloud.
  • Cloud-Native RAN: Breaks functions into containers and microservices that can scale up/down dynamically.

Why It’s Useful

  • Flexibility: Move functions where needed
  • Scalability: Adjust to user demand
  • Efficiency: Use hardware better
  • Automation: Self-healing and self-updating networks

Challenges

  • Latency: Virtual networks can be slower for real-time tasks
  • Performance: May need special accelerators
  • Complexity: Requires powerful orchestration tools       

Why it matters:

Writing about the benefits of virtualization, Sachin Katti of Intel states that "Full virtualization of the RAN . . . offers immense technical and business benefits, including agility, flexibility and scalability. It supports new innovations, like AI algorithms across functions in the RAN, evolving networks to deliver ever-more capabilities at an optimized cost," and enables network operators to "easily implement dynamic power management and network function redistribution. Failures in network operation and system upgrades can be handled by moving the network workload to a different server – without sending technicians to the field"(1).

VNF (Virtual Network Function) → CNF (Cloud-Native Function)

  • VNF: Runs in virtual machines on standard servers (e.g. Linux KVM).
  • CNF: Runs in containers, splitting the VNF into smaller microservices managed by Kubernetes — more efficient and portable
  • Containers are small, fast programs that can be controlled by tools like Kubernetes, which help run more or fewer of them as needed

Importance:
CNFs start up faster, scale out automatically, and use less hardware — all ideal for cloud-native Open RAN.


Sources

Figure 1: https://dgtlinfra.com/network-functions-virtualization-nfv/

1: https://newsroom.intel.com/opinion/future-ran-virtualized-open

Click links below for practice!
Layers of NFV ArchitectureCloud RAN vs. Cloud-Native RANWhy NVF is useful

Sources

  • https://www.o-ran.org
  • https://www.o-ran.org/specifications
  • https://www.o-ran.org/blog/the-o-ran-alliance-security-work-group-continues-defining-o-ran-security-solutions
  • https://www.o-ran.org/press-releases/the-o-ran-alliance-and-the-telecom-infra-project-tip-reach-new-level-of-collaboration-for-open-radio-access-networks
  • https://telecominfraproject.com/openran/
  • https://www.3gpp.org
  • https://www.ngmn.org
  • https://www.etsi.org

Copyright © 2025 PRAXEUM - All Rights Reserved.

Powered by

This website uses cookies.

We use cookies to analyze website traffic and optimize your website experience. By accepting our use of cookies, your data will be aggregated with all other user data.

Accept