By Dr Keshab Panda
Network Function Virtualization (NFV) has the potential to affect significant changes in the way telecom networks are built and operated. It has, in recent times, become a catalyst for major transformational change. Many service providers, telecom equipment manufacturers – all part of an emerging NFV ecosystem – have announced proof of concepts, live trials, and commercial products embracing the NFV vision. Forums such as the ETSI NFV Industry Specification Group and the Open Networking Forum are providing the necessary structures to support its momentum and evolution.
Interestingly, ETSI, in the course of defining use cases and roadmap for implementation of NFV, is now defining Proof of Concepts (PoCs) to test many hypotheses surrounding its various elements, with the active participation of organizations around the world.
While almost every POC has a different set of goals and aims at a specific domain, a detailed study reveals that certain work areas have been commonly covered across multiple POCs. Ten major work areas can be identified (see box).
10 NFV use case work areas
1. Core Network Functions Virtualisation 2. RAN/Base Station Virtualisation 3. Home environment (Set-top-boxes) 4. Content Delivery Networks 5. Service Acceleration 6. Management and Orchestration 7. Charging and Billing 8. QoS/ QoE/ Availability/ Performance Characterization 9. VNF Forwarding Graphs 10. Software Defined Networking
Management and Orchestration is the most common work area. This is concerned with automation of the entire lifecycle management including aspects like instantiation, upgrades and termination of VNFs, data model declaration and usage, VNF health monitoring and repair, elastic scaling of VNFs, virtual machine migration, etc.
Core network functions such as IMS functions, LTE EPC (Long-Term Evolution Evolved Packet Core), Policy Management and S/GiLAN functions form the second most popular work area.
VNF Forwarding Graph and Software Defined Networking (SDN) are the next largest categories, bearing out the fact that operators see value in leveraging the use of SDN in the NFV space. SDN here refers to POCs that use explicit SDN concepts such as separation of the Data and Control Planes, usage of an SDN Controller, etc.
In a virtualized environment where network functions are dynamic and distributed, it is important that service quality is maintained. This is reflected in the significant number of POCs that touch on QoS, QoE and availability/resiliency concepts.
Virtualization of RAN/Base station network functions, charging / billing related functionalities, VNFs in residential devices like set-top-boxes, and the use of hardware accelerators to achieve carrier grade performance are some of the other work areas that the POCs address.
With the rapid growth of rich media applications and the decrease in voice revenues, communication service providers (CSPs) find themselves in a position of having to deal with exploding network traffic while their revenues stagnate. Moreover they face competition from over-the-top players who are agile, flexible, and able to roll out revenue-generating services much faster. NFV applies virtualization to the telecom core network functions and the associated core and value-added applications of CSPs, thereby enabling them to reduce cost and improve time-to-market.
Accelerated adoption of NFV can dramatically reduce both the cost of building their networks and the cost of network operations. NFV enables significant benefits through deployment of virtualized network applications on shared infrastructure, and it brings additional benefits through IT including:
- Reduced Capex and Opex
- Reduced power consumption
- Accelerated time-to-market
- Facilitation of network appliance multi-version and multi-tenancy
- Targeted service introduction
- Rapid scale-up/scale-down of services as required
- Increased innovation through openness
Going forward, NFV will transform the way that CSPs architect and design parts of their networks using standard IT virtualization technologies. NFV will allow them to consolidate many applications and networking appliances onto industry-standard high-volume servers, switches, routers and storage.
As more network functions migrate to standard-IT, high-volume server environments, the stronger the business case becomes for widely deploying cloud and IT datacenter concepts in the telecom network. From a service provider perspective, NFV not only promises Opex savings but also, more importantly, provides a platform for rapid service deployment and monetization, which directly addresses growth models and revenue streams.
For equipment manufacturers, NFV facilitates the development of telecom solutions with optimized performance, software reuse, increased use of open source software, and application scalability in the cloud. On the other hand, NFV will necessitate business model challenges, specifically how the leading network equipment providers charge for their products.
NFV is a disruptive concept, in many ways and driving widespread adoption over the next few years will require significant cooperation and partnerships between service providers and their IT/equipment suppliers.
(The author, Dr Keshab Panda, Chief Executive at L&T Technology Services, has over 27 years of global industry experience in research, conceptualizing, creating, operationalizing and turning around complex IT and engineering services businesses. He is based out of New Jersey in the US)