At a basic level, network virtualization is fundamentally the same whether regardless of where in an operator’s network it is implemented – be it access, edge or core. Briefly, these foundations include:
- The ability to scale-in/scale-out network resources. In order to accommodate and compete with cloud native service providers (think: Netflix, et. al), network operators need the ability to quickly scale network resources in multiple directions. In a nutshell, it refers to the ability to dynamically assign network resources to support new service launches, and the data sessions that come with these launches. Conversely, it also refers to the ability to quickly reclaim these resources when they are either no longer required, or needed elsewhere.
- Enabling service flexibility. As more enterprises move their services to the cloud, the ability to enable a variety of service definitions and QoS parameters also becomes important. To this end, virtualization at the edge promises to help service providers to choose an appropriate access type, and map that to the core transport link that is most appropriate for the particular application performance that is trying to be achieved.
- Managing costs and/or avoiding vendor lock-in. In a physical network functions environment, operators are largely dependent on the product cycles of their suppliers to incorporate customizations and/or introduce new features that help them to either manage unique network design requirements, or meet evolving customer expectations. In a virtual network functions environment, the fact that the elements are largely software-driven and can run on off-the-shelf hardware means that product development can be done in more agile ways. In turn, this helps to enable much faster product development, which helps operators to better cope with rapidly changing networking requirements.
Edge virtualization is where 5G will ‘live or die’
It is often said that 5G is about more than introducing a new wireless access technology into the networking world. Indeed, 5G can be described as creating the ability to match any access technology to the appropriate transport, and core, networking resources to meet a variety of use cases. As we move forward to the 5G era, it will take some level of virtualization at all points in the network (i.e., access, edge and core) in order to provide the required scalability, service agility and cost structure needed to support the variety of use cases that 5G promises to support.
In this regard, edge network virtualization becomes one of, if not, the fundamental 5G enabler. In addition to the need to scale-in and scale-out on an agile, if not completely, automated basis, 5G will require rapid, software-centric product development cycles, and the ability to handle an unprecedented mix of service requirements. It will also require additional technologies such as Multi-Access Edge Compute (MEC) and network slicing that are also rooted in the ability to virtualize at the edge. To this end, while comprehensive network virtualization is vital to 5G, it can be argued that the edge is where 5G will differ most from previous wireless networking generations.
For additional reading on the subject:
- Download: Edge Virtualization: Near-term deployment drivers for 5G service enablement produced by Layne Bridge and Associates for RCR Wireless News
- Edge Virtualization: Near-term deployment drivers
- Edge Virtualization: 5G service enablement
- Edge Virtualization: Overcoming deployment hurdles
- Edge Virtualization: Forging the path forward
