5G Network Slicing: A Golden Opportunity for Telcos?
With telcos already investing heavily in the digital transformation of their traditional, physical, siloed networks, they are also looking for ways to monetize the results with the goal of making profitable return. One of the ways in which they can do this is through network slicing.
In basic terms, network slicing is a means of creating many logical, virtual networks that can operate simultaneously on the same physical architecture. Gaining new networks with capabilities dedicated to different applications “on demand” without having each time to invest in specialized hardware each time sounds like a license to print money – and it could well be just that… if the underlying architecture is optimized for 5G in the first place.
5G Network Slicing in 2020
It’s likely that this year will see 5G rollouts increasing pace globally, which is why telcos have already been investing to ready their networks for this technological breakthrough. Operators know the promise of network slicing, and are eager to reach the promised land sooner rather than later. This will be relatively simple at core network level, as network function virtualization (NFV) and software-defined networking (SDN) are both becoming more commonplace and are natural enablers for such transformation. This is because, once you can instantiate virtual network functions, you can create new network slices by having separate VNFs for each one. However, it’s worth noting that even the most advanced network operators are limiting their ambitions to just 70% network virtualization by the end of the 2020 – so NFV/SDN isn’t going to become so pervasive this year.
The situation is different in the radio access network, where full blown network slicing capabilities required the ability to share a pool of radio resources (RF bandwidth) between slices. The current approach is based on multiple radio access technologies (RATs), which is well known from the coexistence of UMTS, LTE and NB-IoT cells on the same site. But that “conservative” approach limits the number of potential slices, excludes on-demand slice creation, and creates inefficiencies (there is no traffic multiplexing between slices). These limitations originate from the fact that 5G RAN implementation lacks full support for network slicing. This situation may not change until we start to see good business cases for having such capabilities in the 5G RAN network. Nevertheless, the limited approach seems sufficient for 2020.
What are the possibilities and limitations of 5G network slicing for telcos?
The most compelling business use cases for slicing will be connected with low-latency applications. On the other hand, these are just the types of applications that seem most challenging – because the majority of 5G rollouts are still not Standalone 5G (rather, they are NSA-5G). As low-latency applications usually require edge computing capabilities (and therefore full-blown implementation), they are only possible with standalone 5G (SA 5G). In the absence of SA 5G, the proprietary mobile edge Compton (MEC) solution may be required, as non-standalone 5G architecture will remain predominant in 2020. This may hinder they deployment of scalable solutions.
In conclusion, network slicing will progress in 2020. But, as operators adopt the “low-hanging fruit” strategy, it’s likely that we will have to wait for full implementations of network slicking in RAN and core networks.