In this article

Network function virtualization (NFV) describes the use of technology to virtualize network services, including load balancers, firewalls, routers and switches that traditionally run on proprietary hardware to reduce costs and make them faster and more agile.

Essentially, it's a way to accelerate service deployment and reduce operating costs of network operators by decoupling network functions from dedicated hardware and migrating them to virtual servers.

This enables service providers (especially telcos) to run their networks on standard servers instead of proprietary ones. Leveraging virtual machines also removes the need to use dedicated hardware for every network function.

This means service providers can purchase inexpensive servers, storage and switches to run virtual machines that execute network functions rather than install proprietary hardware. As such, multiple network functions can be handled by a single physical server, which minimizes truck rolls and reduces operating/maintenance costs.

What are the core benefits of Network Function Virtualization (NFV)?

The two main benefits of NFV are improved agility and scalability. With this model, service providers no longer need to requisition additional hardware resources to dynamically add capacity and deliver new applications and network services. 

For instance, there's no need to deploy new hardware across the network to enable encryption. Admins can deploy encryption software on a switch or standardized server that's already in the network. Also, service providers can simply create a new virtual machine whenever there's a need to add a new network function. 

Combined, among other benefits, these factors help greatly improve agility and scalability.

What are the components of NFV architecture?

The primary goal of network function virtualization architecture is to transform how operators and service providers build networks by using IT virtualization technology to deliver and consolidate network functions via industry-standard servers, storage and switches. 

A typical network function virtualization architecture consists of the following layers:

Virtualized network functions (VNFs)

Virtualized network functions are software applications used to deliver a range of network functions such as IP configuration, directory services, mobile core, routing, security and file sharing.

Network functions virtualization infrastructure

Network functions virtualization infrastructure (NFVI) delivers several infrastructure components, including networking, storage and compute on a platform such as a container or a hypervisor that supports software that can be used to run network apps. Essentially, it's a hardware and infrastructure software platform that is needed to run network applications.

Management, automation and network orchestration

This describes a framework for provisioning, managing and orchestrating new VNFs and NFV infrastructure.

What are the top NFV deployment options?

Network operators can configure their NFV deployments in several ways through a combination of suitable applications, orchestration, virtualization infrastructure management (VIM) and hardware platforms.

VNFs are the application layer of NFVs. They provide the software code that delivers feature-rich network applications. Multiple VNFs can be service-chained to deliver broad network functions for sophisticated use cases. There are dozens of proprietary and open source VNF options available to network operators.

A critical part of every NFV deployment is reliable, high-performance servers. Since network function virtualization relies on server virtualization technology, operators can choose any of the popular hypervisor options, such as VMware and OpenStack. Containers are also an alternative option, although they are not widely used in network function virtualization deployments.

There are several open standard or vendor-based MANO architectures available to network operators. The Open Network Automation Platform (ONAP) is one especially popular open-source option.

Why should you use NFV?

At its base level, network function virtualization offloads networking functions onto software that can run on widely available industry hardware, thus reducing the need for service providers to use expensive, proprietary hardware. Separating network function from dedicated hardware delivers several benefits including:

  • Decreased maintenance and hardware costs.
  • A longer lifecycle for your network hardware.
  • Easier network upgrades.
  • Reduced network maintenance costs.
  • Minimizes network power consumption.
  • Less space needed for network hardware.

By enabling enterprises to modernize their networks with modular software running on standard server platforms, NFV enables high-performance networks with greater adaptability, elasticity and scalability. It's suitable for a wide range of applications including 5G, Internet of things and SD-WAN, among others.

Other top network function virtualization applications include:

  • A range of security functions including NAT, intrusion detection and prevention systems, firewalls and more.
  • Network monitoring
  • Network slicing
  • Content delivery networks (CDN)
  • Virtual customer premises equipment (vCPE)
  • Video servers
  • Session border control (SBC)
  • IP multi-media Subsystem (IMS)
  • Evolved packet core (EPC)

Learning the value of NFV

Service providers are gradually modernizing their networks and moving services and workloads out of data centers to virtual machines and the edge of the network. From there, they can deliver more efficient services with higher bandwidth and lower latencies.

With network function virtualization, service providers no longer have to run network functions on dedicated hardware. Furthermore, the virtualization of network functions enable a single server that can handle multiple functions, which means providers use less hardware resulting in cost reductions due to less physical space, reduced running costs, power consumption and maintenance overhead.