By: Karel Hodan

Today, virtual networking is the new paradigm for building networks, and it is critical to the success of 5G. In essence, it leads to a decoupling of physical infrastructure from the services that the network delivers.

Yet, virtual networking contains an inherent paradox. With virtual resources, service level management is completely separated from the underlying physical resources required for service delivery. The capacity required for a service is drawn from the capacity of the virtual resource, which is configurable at the level of the physical resource. But while the network core and other infrastructure may exist as virtual assets, a service is delivered to something—a device, an object or an individual.

In this article, we’ll examine the paradox and discuss how service providers and telecoms operators can resolve the problem, enabling them to efficiently and effectively deliver dynamic, virtual resources.

First, a look back

To understand this paradigm shift, let’s start by looking at how networks have historically been built. Generally, resources and services were tied to hardware. This approach meant that networks typically required specialist, proprietary hardware, which created lock-in, limited choice and tended to contribute to interworking issues. The growth of standards-based commodity hardware addressed this situation by providing common physical elements that could be reused for multiple use cases, but services and elements were still tied to specific resource clusters.

By and large, virtual networking is perceived as the answer to this problem. It enables resources to be created dynamically in software to reflect changes in service demands, with the support of data centers composed of commodity hardware blades and processors. If, for instance, more capacity is required for a service, it is allocated at the level of the virtual resource. When more physical capacity to deliver the virtual resource is required, additional hardware can be allocated, and vice versa.

According to SDX Central, virtual networks offer promise by enabling “the service provider network to provision the most suitable and efficient networking structure for the applications it hosts—and to alter that structure as conditions warrant, using software rather than requiring physical changes in connections to hardware.”

As a result, virtual networks offer elastic flexibility. They are also fundamental to 5G, because 5G provides a completely converged network and service architecture, allowing different forms of access connectivity. Today, though, service providers are at a transition point: many providers have hybrid networks, spanning legacy physical infrastructure, as well as new virtualized elements and services. As we move to 5G and into the future, transitioning to virtual networks is key to success.

Service delivery

Networks deliver services that are consumed by people and devices. This means that if, for example, an enterprise requires more capacity to meet its evolving needs, the additional capacity will be provisioned by its service provider and will be made available across the physical path to the location of the enterprise.

If differentiation in terms of specific services delivered across that capacity is required, the service provider will make appropriate allowances. The same is true for wireless connectivity. In short, the provider can shape the services and capacity available in each cell site and across the network.

5G changes the game

Traditionally, networks have been optimized for voice, data and messaging services. However, 5G changes things entirely. 5G networks—wireless or fiber access—are optimized for several different parallel services. These are: Enhanced Mobile Broadband (eMBB), Ultra-Reliable Low Latency Communications (URLCC), Massive IoT, and Fixed Wireless Broadband Access (FWA).