Centralized Vs. Distributed vCPEs: What’s the Best Solution for Your Business?

NFV - Part 2 Are NFV & SDN Both Key to Driving the Fourth Industrial Revolution

Software-defined networking (SDN) and network function virtualization (NFV) are currently two of the most exciting areas of technological innovation as their influence is now being felt around the world, creating a shift towards virtual, software-driven solutions and away from physical hardware. As this transition is occurring, virtual customer premise equipment (vCPE) has grown into one of the largest use cases of on-site enterprise applications.


According to a 2016 survey by iHS Markit, the dollar value of business opportunities associated with vCPE are forecast to go beyond $1.5 billion worldwide by 2020 as service providers try to leverage the cost benefits of vCPE in order to build their SDN and NFV applications.

With more and more businesses and organizations looking to drive down costs and improve their business and services through the use of software-defined networking and network function virtualization, virtual CPEs are becoming an increasingly popular way of delivering network services and applications such as routing, firewalls and virtual private network (VPN) functions.

In this article, we’ll be looking at how best to integrate virtual CPEs into your business through either a centralized or distributed framework. We’ll also be going over the differences between the two as well as taking a look at how these two methods are used in combination within some hybrid frameworks.

So, let’s get started.

Centralized vCPEs

Centralized virtual CPEs are located within the data center and are mainly used by larger businesses an enterprises looking to begin to invest in and integrate software-defined network solutions and network function virtualization.

Within a centralized virtual CPE framework, vCPE is used as a virtualized network function and should be integrated on carrier-grade computing hardware appliances within a service provider’s data centers. The computing appliance is then able to run as router, firewall, VPN and other network services as well as edge routing and unified threat management (UTM).

This process creates a virtual network on top of the physical network and allows service providers to expand and accelerate development and innovation of services, as well as other network requirements such as provisioning. Centralized vCPEs will also work well with other integrated solutions.

Orchestration engines can also prove to be an extremely worthwhile investment for those looking to invest in centralized NFV, however, when considering a centralized solution, the following features are widely regarded as critical; Centralized policy automation, Centralized hybrid WAN management, Network wide visibility and segmentation, Public key infrastructure (PKI) certificate management, Plug and play from day 0.

Distributed vCPEs

In a distributed virtual CPE framework, the vCPEs are deployed depending on the customer’s premises and placed in locations that best suit their function. For example, certain communication service functions NFV infrastructure is better distributed in either a geographically or hierarchically appropriate manner in order to optimize both resources as well as efficiency.

One of the biggest limitations to a centralized virtual CPE framework is the limitations in bandwidth. In today’s network traffic, rather than just being voice or photo, we know have video and audio data that may come in a significantly larger file size than we have seen in the past.

As we continue on through the Fourth Industrial Revolution and the integration of an ever increasing number of smart and IoT technologies, the size and volume of this data will also likely increase.

This is where distributed vCPEs come into their own and offer benefits centralized frameworks are unable to.

A distributed, multi-hierarchy virtual CPE placed based upon the customer’s premises is likely to benefit those companies and organizations who are looking for features such as network offloading during streaming, increasingly distributed authentication, and traffic routing to a much closer peer point. This would also help to reduce latency and improve bandwidth.

Which is Best for You?

Choosing between a centralized and distributed virtual CPE framework is most definitely dictated by your individual circumstance and resources. If you are part of a larger company who already has data center resources and is looking to improve upon or better manage your SDN or NFV offerings, then a centralized vCPE may be worth looking at.

If, however, you find you do not have the resources or capital required for such a transition, a distributed vCPE framework may be much more suited to your requirements as it would also allow you to scale up or down depending on what was required. This is a much more convenient solution for smaller businesses and organizations who aren’t as qualified in such areas just yet.

In order to enhance and improve their software-defined networking operations, consumers look to network virtualization to decouple their network functions such as DNS, caching, IDS, and firewalling from the proprietary hardware that were, until recently, the dominant solutions, so as to enable them to run on software instead.

It is hoped that this will lead to improved operations and services for both customers and service providers through the use of more advanced architectures and more efficient frameworks.


Related Posts