With the freezing of 3GPP release 16, the deployments of 5G New Radio (NR) and the anticipated launch of Apple’s 5G iPhone, all that’s needed to make 2020 the year of “real 5G” are deployments of the 5G standalone core.
That step, as a substantial inflection point in the chronology of 5G, should not be underestimated. It will either shape the real commercial success of 5G or could merely relegate it to the status of “just another G.” A combination of strategy and technology will determine the delta between the two.
From a strategy perspective, approaching 5G from a “build the services that need the network” (more on that later) approach is crucial. From a technology perspective, it is about integrating a converged charging system (CCS) that is capable of monetizing digital-first oriented services at a virtual scale against demanding service level agreements (SLAs). Achieving this across consumer, enterprise, industry, private and IoT segments, while encompassing progressive new business models such as B2B2X, is paramount.
The work undertaken by many telcos across the globe has made it clear that even the most advanced and progressive will be using 2020 as a “sandbox” year. It will be a proof of concept time for pure 5G core with 2021 earmarked for live new customer deployments and the beginning of the migration from 4G to 5G.
Following early debates around 3GPP defined options 3, 4 and 7 NSA (non-standalone) core deployment, and options 1,2, and 5 SA (standalone), the world appears to be bi-furcating around two clear options. Option 3 is for those deploying 5G new radio but wanting to control it and the 4G E-UTRAN from an Evolved Packet Core (EPC). Those wishing to push ahead with option 2 deployments manifests as a pure 5G core controlling a 5G new radio access network.
The above matters because decisions made in 2020 by leading telcos will have a profound impact on the future relevancy and role of this great industry.
We are now some 20 years into the mobile broadband era (yes 1G and 2G came earlier but were not geared to providing broadband data services) and in that time, we have witnessed the “data connectivity era” of 3G and the “capacity era” of 4G. Both, it could be argued, were built on the premise of building the network that needs services with an underlying operational approach of a high percentage of revenue being invested into network CAPEX.
Clearly, that network-centric approach, while generally driving excellent mobile broadband, has not resulted in the sustainable, differentiated revenue and margin growth that telcos need to maintain relevancy. Connectivity alone has become a commodity.
Services That Need the Network vs. Network That Needs Services
If 5G is to change anything, it will only be through a fundamentally disruptive “build the services that need the network” philosophy. It needs an approach where an increasing percentage of revenue is invested in service R&D, balanced against lowered network CAPEX investments. Given that a high percentage of mobile network costs sit in the RAN, activities such as passive or active RAN sharing, OpenRan and dynamic spectrum allocation may somewhat reduce individual MNO CAPEX loading.
The key point here is that while having an option 3 bridging strategy between 4G and 5G makes some sense, it also perpetuates the “build the network that needs services” philosophy and looks at the world through a connectivity lens. Option 2 5G core provides the foundation for a progressive “build the services that need the network” approach that is the basis of a profound business and relevancy transformation for the telco industry.
The Impact of the 5G Core
The evolution of the mobile core over the last 20 years could best be summarized as going from monolithic to micro. The monolithic architecture of the 3G GPRS core network, based around the Gateway GPRS Support Node (GGSN) and Serving GPRS Support Node (SGSN), evolved into a slightly less monolithic LTE Evolved Packet Core (EPC). That was founded on the Packet Data Gateway (PGW), Signaling Gateway (SGW) and Mobile Management Entity (MME), all part of the System Architecture Evolution (SAE). This period also saw the release (in 3GPP release 14) of the principle of Control and User Plane Separation, or CUPS, which aids the virtualization and associated flexibility and scalability of both the PGW (became PGW-U and PGW-C) and the SGW (became the SGW-U and SGW-C).
The arrival of 5G core in 3GPP release 15 saw a deeper push towards a virtualized, micro-service design with the founding principles that it should:
Be based on a service-based architecture (SBA)
Utilize web-scale internet protocols such as HTTPv2 and move away from telco protocols such as Diameter
Interact on a “publish and discover” principle allowing for greater scaling of functions and the interaction between them
Have separation of the control and user planes (CUPS)
Be independent of access technology in support of wired and wireless convergence
Have open access to third-party interaction
Have a “micro-service” construct of functions, dedicated to specific tasks and scaled via virtualization schemes, such as NFV, Kubernetes and Docker
3GPP defined a “micro-service” approach as encompassing the following key functions (not a complete list). Each function has a tightly defined, specific role and can scale independently of the other functions:
AMF – Access and Mobility Management Function
AUSF – Authentication Server Function
SMF – Session Management Function
UPF – User Plane function
UDF – Unified Data Repository
UDM – Unified Data Management
NRF – Network Repository Function
NEF – Network Exposure Function
N3IWF – Non-3GPP Access Interworking Function
NSSF – Network Slice Selection Function
NWDAF – Network Data Analytics Function
PCF – Policy and Control Function
CHF – Charging Function (5GC Network Function Residing within the CCS)
CCS – Converged Charging System
In a case of the “whole being greater than the sum of the parts,” the real benefit of the service-based architecture approach lies in its ability to support a wide array of UEs or end devices at scale. Whether consumer, enterprise, industry or IoT, it ties to the ability to deliver a rich combination of homegrown (possibly IMS based) and third-party B2B2X services via the Network Exposure Function (NEF).
Additionally, the opportunity is there to virtualize service delivery enabled through the Network Slice Selection Function (NSSF), selecting the appropriate slice for the user/service and to then police and account for usage and event activity via the policy control and converged charging system.
5G Won’t Replace 4G Overnight
In building the new 5G core, telcos will need to work with vendors who can offer best in class capabilities that provide a bi-modal bridge between both worlds. In the case of MATRIXX, which delivers rich CCS functionality in 5G and OCS functionality in 4G (and support for previous Gs), “best in class” has a clear meaning. It means delivering a real-time decision engine capable of executing all commercial business logic in one place across all business lines and models (consumer, enterprise, industry, private, IoT, B2B2X). It achieves this by executing the industry’s highest commercial transaction performance rate per unit of footprint at a consistent <10mS transaction latency to meet the expected four-fold increase in transaction load that a combined CCS/OCS will need to handle during the 4G to 5G transition period.
The commoditization clock is ticking. The need to adapt to a “build the services that need the network” approach is crucial for the mid-term health and relevancy of telcos. The acceleration of 5G core build and deployment is essential.