5G's raison d'ętre is to quickly provision real-time services and applications across end-to-end networks to customers and business in various locations. While that seems to be a tall order, the technologies are coalescing to deliver on the 5G promise of faster, better and new services.
Here's a look at some of the technologies and other considerations for the deployment of 5G services and applications.
The CIO view
While 5G applications are currently in the pilot stages, their adoption is expected to accelerate as a result of the demonstration effect from the 2018 Winter Olympics early next year in South Korea. Given the fluidity of the situation of 5G -- especially in pre-standards deployments -- how can CIOs plan their budget allocations in the medium-term future?
Barry Graham, director of product management, at the TM Forum, asserts that it's not all about the future. The industry has prioritized and enhanced the broadband aspect of 5G, which is more familiar to service providers. Mission-critical applications such as autonomous cars are further away on the horizon while massive machine-type communications are even farther down the road, but the market size for the latter is expected to be in trillions of dollars. When its deployed, machine communications will be spread over hundreds of potential applications and the customized versions will be attractive because of the prospect of higher profitability.
"Internet of Everything applications are characterized by heterogeneity and a diversity of platforms," Graham said. "They will benefit the most from a robust and pervasive cloud-native environment that can scale up and down."
For a pervasive cloud-native environment, "the industry needs standards for the communication between the edge and the mobile edge when traffic does not have to be directed to the central cloud," according to Graham.
"While making the best of their existing investments, CIOs should begin to plan for the Internet of Everything for use cases such as micropayments," he concluded.
The edge cloud, enlarged with the incorporation of a cloudified RAN, will be transformed to meet the individual performance needs of 5G-enabled applications. Services can be tailor-made for customers and delivered in real-time by placing all or most of the elements for service composition -- such as VNFs, virtualized resources, microservices, management and orchestration software, a cloud-native infrastructure that includes the SaaS. IaaS, PaaS, and Cloud-RAN -- in close proximity to customers at the edge.
The unknown at the moment is the economic feasibility of the edge clouds. "Clouds are centralized for a reason; their economic returns are known. For edge clouds, network operators and the industry must now work hand-in-hand to develop financially feasible use-cases together," said Franz Seiser, vice president of Core Network and Services at Deutsche Telekom.
A step change in the expected performance of 5G undergirds the confidence to adapt to market changes as they happen. Compared to the latency of 15-20 milliseconds for 4G LTE, 5G will be below 4 milliseconds for broadband and as low as 0.5 milliseconds for mission-critical applications. 5G has actual throughput speeds of 500 Megabits per second to 5 G/bits while 4G LTE's throughput ranges from 6.5 M/bits to 26.3 M/bits.
Service providers as master orchestrators
Service providers will play a pivotal role in orchestrating all of the elements needed for customizing services, which they will source from the stakeholders in ecoystem at the edge of their networks.
The managed service provider's role is build an end-to-end virtual network slice based on the needs of the customer. The managed service provider is brokering the resources from one entity to another while also insuring that service level agreements are being met.
While network slicing has vast potential for enabling specific services and applications on the fly, service providers need to insure that they have insight into the end-to-end framework including inter-slice management.
"ONAP is one potential candidate for a framework to manage cloudified networks, including network slicing," Duetsche Telekom's Seiser said.
Network slicing provides the means to customize business solutions for verticals based on the performance needs spelled out by the SLAs signed with customers.
"5G system aims to provide a flexible platform to enable new business cases and models... network slicing emerges as a promising future-proof framework to adhere by the technological and business needs of different industries," according to a European Union document.
The keystone of network slicing and its design challenge is "network slicing needs to be designed from an end-to-end perspective spanning over several technical domains (e.g., core, transport, and access networks) and administrative domains (e.g., different mobile network operators) including management and orchestration plane," according to the European Union document.
On the stretch from the core to the access networks, each operator is sharing network capacity with peers or multiple tenants for the delivery of one or more of the portfolio of services, which saves on capital expenditures that need to be minimized in order to make investments on customized services more viable.
Industrial automation, such as collaborative robotics -- which includes humans interacting with robots -- is a vertical where network slicing is expected to gain acceptance.
"The latency and throughput requirements of collaborative robotics vary within a factory floor, or across an industrial region, and network slicing flexibly tailors network services for each of them" said Harbor Research analyst Daniel Intolubbe-Chmil.
Network slicing end-to-end -- from the core to customers' facilities -- will become possible with the reliability, flexibility and the desired throughput (collectively described as network determinism) that cloudified RAN networks will be able to deliver.
"In conjunction with mmWave technologies, they will support throughput comparable to Ethernet. Beamforming helps to achieve high reliability that maintains the quality of the signal end-to-end. Small cells lend greater flexibility and speed in deployment compared to the Ethernet," Intolubbe-Chmil said.
SK Telecom deployed its first cloud RAN in the third quarter of last year in order to support a rollout of LTE-A, which is pre 5G, while also preparing a base for its upcoming 5G platform, according to Neil Shah, research director at Counterpoint Research.
"Network operators are ready to scale cloud RANs as they have greater clarity on the right mix of macro cell sites and small cells controlled by cloud baseband units," Shah said.
Virtual reality looks to be the leading application of choice in the 5G environment. The countdown for its mass use has started as the organizers of the 2018 Winter Olympics next year in South Korea engineer their stadiums for 5G-enabled virtual reality even before the 5G standards are finalized.
Sports fans will be delighted with 360 degrees virtual reality that includes the option to view footage from their preferred angle such as from a sportsperson's head-worn camera or a drone with a more panoramic view. They will also be able to switch streams from one sports arena to another with virtual reality.
"An edge cloud placed in the sports stadium, meeting the processing demand from VR streams flowing to hundreds of fans, lowers latency," Seiser said. "Bandwidth is used efficiently when the data rate is reduced by delivering only a sliver of a VR stream needed to render content for the field of view of the user," Seiser explained.
With a constellation of services and applications at the mobile edge, 5G cloud-native architecture are starting to converge to make widespread customization of services possible for service providers and their end customers.
— Kishore Jethanandani, Contributing Writer, Telco Transformation