Originally published in Wireless Week on October 28, 2016
While much has been written and speculated about the next radio standard that is destined to form the basis of the forthcoming 5G standard—dubbed 5G NR for “new radio”—the core network is also a vital piece in achieving the goals set forth by the standards committees, and in helping to assure the competitiveness and relevance of network operations for the coming decades.
Over-the-top service providers like Facebook, Microsoft, and Apple do not (yet) have any access networks, wired or wireless. Google has been building its own fiber optic and WiFi networks over the past several years, but currently these companies rely solely on their data centers (or network cores) to provide their services.
Not only is an operators’ core network fundamental in how they manage and route all of their data voice and video traffic, it is the engine with which they charge their customers, house their applications and content, and is the point where they communicate with other operators and Internet properties. The core network is where operators can create the services, plans, and content with which to differentiate.
An operator’s strategy for the radio access network is typically to have at least two vendors for competition and, because it is so capital intensive, to drive the costs as low as possible. Operators also work with the standards groups to commoditize the products as much as possible. While there are similar standardization and commoditization efforts in core networks, and improved economics and flexibility in the core network is very important, there are much more areas for competition and differentiation.
There are five large, global radio base station vendors: Huawei, Ericsson, Nokia, ZTE, and Samsung. The top three account for perhaps 90 percent of the market and this is not expected to change with 5G. Those vendors are able to invest in the research and development necessary to develop and produce such a complicated and important set of standards. Therefore, the barriers to entry are very high and the profit margins are relatively low.
Conversely, barriers to enter the core network, since it is becoming software-only, are falling. The increasing reliance on open source software as the basis of much of the software frameworks is also lowering barriers to entry and inciting innovation.
5G is the first set of wireless standards that is being designed without the smartphone at the center. Therefore, network operators must consider a wide variety of applications and traffic profiles.
5G networks, using a common underlying software infrastructure, must be able to quickly adapt to these disparate applications and traffic patterns. It also must be able to “breath,” to expand or contract based on split-second changes in demand. Each application will have a unique network profile that will be defined as a set of services within the core network, these profiles are often called network slices.
Network slicing, and the ability to dynamically and automatically expand and contract the capacity of them, is a key element of 5G standardization and will incorporate or rely upon the following technologies:
Network Function Virtualization (NFV) & Containers
The ability to use low-cost generic hardware as the basis of networking infrastructure is a key foundation of 5G standards, as network operators begin to re-architect their core networks for virtualization and/or containerization. Virtualization is the beginning step of making core network functions more economical, scalable, flexible, and open.
Management and Orchestration (MANO)
Management and Orchestration takes the idea of virtualization a large step further and enables the various network functions to work together (i.e. service chaining) and allows for automated changes in network capacity.
Machine Learning and Artificial Intelligence
Machine Learning and Artificial Intelligence are key technologies for service providers in several ways. As much of the network and customer data from an operators’ disparate networking, administration, operations, and billing systems gets aggregated into Big Data databases, ML and AI systems can analyze the data for: (1) advertising and revenue generating opportunities, (2) security and fraud attempts, (3) ways in which to optimize and plan for future network resources, including increasing and decreasing network capacities at a very granular level.
Micro-services represent the disaggregation of current network elements into more fundamental network functions, like “control” and “data path”. These micro-services can be scaled independently, increasing network performance and improving efficiency.
Tacticle Internet refers to the type of experience that can be achieved with extremely low latency and very high speeds (one-millisecond latency of 5G) necessary for applications like autonomous cars, virtual reality (VR) or augmented reality (AR)
To achieve these latency times, key network functions must be pushed as close to the user as possible. Mobile Edge Computing represents this trend. Servers will need to be placed well outside of central offices/data centers. They may need to be incorporated into the basestation or tower equipment itself, while traditional basestation functions, like baseband processing, are moved to a more central location within the network.
5G is much more than a new radio access network and more spectrum. It includes core network architectures and technologies that are akin to what many of the leading OTT service providers are using today. These new technologies will allow traditional operators to define new services for a vast array of devices and services, some with nearly instantaneous response times and extremely high bandwidth (autonomous cars), to services with extremely low bandwidth but very long battery life (remote sensors). 5G core networks will form the foundation of the progressive digitalization, marking a new industrial phase.
It’s anticipated that 5G networks and services will begin to be operational in approximately 2020, but the preparations and investments in the foundational technologies are well at hand.