Network Slicing: A Cornerstone of 5G

By Mika Skarp

There are three corner stones of 5G networks. The first two relate to requirements of what 5G will have to deliver, and these include: 1) A 10 Gbps air interface, and 2) A less than 1 ms delay over the network. The third and possibly fourth cornerstones of 5G relate to how these demands will be met and they are largely defined within the concept of Network Virtualization.

Network Slicing represents the primary vehicle for this. And while it will live alongside other network virtualization functions, it is so critical to 5G that it may well itself deserve its very own cornerstone.

The essential advantage of virtual, cloud-based technology is that it makes it possible to scale up and down network elements as needed. But what does this mean, and why is it important?

In a virtualized world, hardware is theoretically "all purpose" and thus comes online at a significantly lower cost than dedicated, 'actual', physical hardware. Of course the demand for network capacity varies a great deal, and may fall and spike over time and across locations. Virtualized hardware allows capacity to move when and where the demand is greatest, and do so instantly and virtually without cost. 

In case it wasn't already apparent, this radically changes the way networks operate and in much the same way that worldwide software licensing models were born at the dawn of digital.

While in our new virtualized paradigm CAPEX is an immediate positive impact, but this is not the kind of disruption we are talking about. Likewise the savings from using less electricity triggered by idle capacity provides a nice footnote, but isn't what this play is about. Nay, the big opportunity lies in network slicing and the very DNA-level change it brings to mobile networks. Now, for the first time, we have a mobile network paradigm that effectively and irrevocably separates the service layer from the infrastructure layer. While this may excite heads of finance, imagine what its doing for architects, engineers, marketers and entrepreneurs?

Layer separation will not only provide a much richer service portfolio at radically diminished cost, but it will liberate innovation to invoke almost 'free play' within the network to roll out highly differentiated new services. No longer will new network features require standardization or changes to the underlying architecture, but indeed it give rise to a new generation of application developers free to dream and invent without restrictions.

Anyone who's been in the telecom business from either side of the technology / marketing fence knows the challenges of building out new services. Up until now simply adding a new feature to a mobile network might take years. In Network Slicing different use cases can utilize different network features and scale in any direction.  Just imagine configuring a new mobile network and to have it up and running in seconds, in about the same time it takes to configure a virtual server. I love 5G!!!