Network slicing first appeared with 4G networks and enabled new possibilities – and this will go much farther when 5G enables network slices to incorporate the radios.
With great power, however, comes great responsibility for getting the quality of service right, otherwise end-to-end services and mobile devices could be in deep trouble.
The power of 5G networks isn’t in its capacity alone — although 5G networks will have the potential for greater capacity at the mobile device and in mobile backhaul.
The power of 5G networks won’t be the actual perceived improvement in the user experience, although that also will be significant. The real benefit is that network slicing will enable application designers and network architects to build end-to-end virtual networks tailored to their applications’ requirements – and implement throughout the entire network, says Marc-Antoine Boutin, director of Product Management at CENX.
Network slicing, while a relatively new idea, isn’t totally new to the forthcoming 5G standards. Elements of network slicing appeared in some 4G architectures, and could be implemented to encompass everything from the data centre to the network core to the network edge.
The radio part of the wireless mobility, however, wasn’t programmable in the network slice – specifically, 4G network slicing left out the RAN (radio area network) and access vendors’ tower microwave links and cellular data networks.
With 5G comes broader completion of the network slicing concept, and this unlocks tremendous possibilities. A network slice still originates at the data centre, and now extends through the core IP/MPLS networks, macro cells and small cells. Network slices can encompasses the cloud RAN, and also program the last-mile wireless connections for mobile users. In other words, it’s got everything.
Slicing and Dicing
Driving the architecture of 5G network slices is software defined networks (SDN), which provides the programmable infrastructure that separates the data plane from the control plane of the network. SDN enables application designers and service providers to define the exact networks they need, encompassing specific data centre servers, cloud providers, network core, backhaul, and front haul services.
Running on top of those applications are virtual network functions (VNFs), specific programs that offer necessary functionality running on the SDN-capable devices via network functions virtualisation (NFV). Lifecycle service orchestration will ensure that the network is configured properly end-to-end, even across multiple carriers, and that the proper VNFs are installed and running correctly.
In other words, 5G network slicing allows service providers and application developers to create the ideal virtual network to tie together their back-end servers, mobile end-points, analytics, firewalls, authentication servers, encryption, revenue systems, and other infrastructure – and layer it on top of whatever physical network and service providers happens to be the ones offering the L1/L2 connectivity.
Part 2 of this article continues tomorrow.
The author of this blog is Marc-Antoine Boutin, director of Product Management at CENX
Comment on this article below or via Twitter: @ VanillaPlus OR @jcvplus
