SK Telecom has been one of the most aggressive carriers when it comes to advancing 5G, as the South Korean operator hopes to have an initial drop of the standard ready to trial by 2018.

Ahead of this month's T Dev Forum, SK Telecom (Nasdaq: SKM)'s developers conference with 400 participants looking to build for a 5G future, Light Reading caught up over email with two of the carrier's key 5G visionaries, Dr. Changsoon Choi and Dr. Jong-Han Park, both of whom are senior managers of SK Telecom's 5G Tech Lab and Corporate R&D Center. (See 5G Use Cases, Pre-Standards Groups Proliferate , SK Telecom, Samsung Team Up on 5G R&D and SK Telecom, Ericsson Collaborate on 5G Research.)

They shared their thoughts on everything from what 5G will entail to how small cells, WiFi, SDN and NFV factor in to meeting the challenges this next-generation presents. Read on for the full interview.

Figure 1: 5G Robotics SK Telecom's 5G Robot debuted at Mobile World Congress in February, showcasing how it will be able to move in real-time to reflect human movements through a 5G-enabled connected network, demonstrating ultra-low latency – a key feature of 5G.

Want to know more about 5G? This will be one of the many topics covered at Light Reading's second Big Telecom Event on June 9-10 in Chicago. Get yourself registered today or get left behind!

Light Reading: There's been a lot of talk about 4G network advancements, 4.5G or 5G-ready technology, but how does SK Telecom define true "5G"?

SK Telecom: SK Telecom holistically defines 5G as end-to-end network architectures and services, promising the greater values in terms of (1) user experience, (2) connectivity, (3) reliability, (4) efficiency, and (5) intelligence. Out of the five values, the first three are the most significant.

5G guarantees a consistent user experience. To support the timely delivery of real-time media with ultra-high quality, one of our goals is to guarantee 1Gbit/s bandwidth per user with 1 ms radio latency.

5G efficiently supports massive connectivity. We expect that there will be diverse vertical services based on IoT, and 5G networks intelligently manage resources in a cost-efficient manner based on the actual usage patterns and requirements of different services.

5G is highly reliable. Being reliable all the time for all the use cases may be prohibitively expensive. Therefore, it becomes important to carefully classify and define certain use cases, such as public-safety networks and selectively provide high degree of reliability.

LR: Will 5G be a "rip and replacement" new network or just incremental improvements to 4G? Which makes more sense from a technology and business model perspective?

SK: 5G is based on both continued evolution from 4G and a completely new, revolutionary technology.

We will continue to develop and commercialize 4G technologies as we move towards 5G, which is expected to launch in 2020. Also, as revolutionary technologies are not available yet, mobile operators need to start thinking about technologies that require incremental improvement, and those that need to be newly developed. Although these revolutionary technologies will not become available all at once, 5G will be a great opportunity for operators to carefully rethink and redesign their networks as many network components will either be enhanced or replaced.

The same thing applies from a business model perspective. There will be new services that can be provided on the evolved 4G. However, these services will be incremental and add slightly increased values. When the new technologies are fully implemented around 2020, a set of new services and business models will also be introduced.

LR: Is 5G an architecture or a new air-interface?

SK: 5G and its services introduce widely varying challenges and requirements. Some of these challenges such as peak data rate may be tackled by a new air-interface, and some challenges such as reduced end-to-end latency may be addressed at an architectural level. In other words, evolutions and revolutions are needed in both the architecture as well as the air-interface to meet the requirements introduced by the services.

More specifically in the 5G vision of SK Telecom, 5G consists of three layers: 1) hyper-connected radio, 2) all-IT infrastructure as enabling platform and 3) innovative services. Several tens of Gbit/s speed and extremely low latency will be provided over the new 5G radio. A service-oriented architecture must support various 5G services by intelligently connecting and orchestrating seemingly disjointed underlying resources. Taking full advantage of the underlying service-oriented architecture, 5G services are expected to offer a highly reliable and immersive user experience. To achieve successful 5G commercialization, all three components are indispensable.

Next page: SDN/NFV, small cells, WiFi and a timeline

LR: How do SDN and NFV come into play for 5G?

SK: 5G introduces many new services with widely varying characteristics and requirements. Therefore, the network has to be sufficiently programmable in order to satisfy the demands made by these services in a flexible way. SDN and NFV will continue to be important technologies in the transformation of the existing network to a more programmable and open network. Internally, from the operator perspective, the programmability will help the operators implement and manage the diverse 5G services in a cost-effective and intelligent manner. Externally, from the viewpoint of other service providers (e.g., OTT), the rich network-asset information and controllability will be used to further improve quality-of-service and also facilitate the introduction of innovative 5G services.

LR: What role will small cells and WiFi play in 5G?

SK: Initial deployment of 5G radio will likely be a HetNet (heterogeneous network), where many small cells are overlaid over a macro-cell. The macro-cell will still be based on LTE-A evolution with a higher degree of carrier aggregation and more antennas. For hotspot areas, new radio access techniques based on higher frequency such as centimeter-wave and millimeter-wave will be deployed to boost wireless capacity.

Since higher frequency suffers from higher atmospheric loss and larger shadowing, more small cells are required to provide similar cell coverage. Thus, the capex and opex reduction of small cells become important. So, WiFi technologies are expected to be one of viable solutions for increasing cellular capacity in a cost-effective manner.

LR: How do you plan to meet the timeline of commercial deployments by 2020? Why is this so important?

SK: SK Telecom has already announced plans to carry out a pre-5G trial in 2018 and to start commercialization, hopefully in 2020.

Considering the timeline of 3GPP, ITU and other standardization bodies, it is likely that 5G architecture and key enabling technologies will be defined first around 2018. SK Telecom has also participated in 5G standardization activities such as the 3GPP SA1 study program. At the same time, we have been actively working with a number of network and IT manufacturers in order to build 5G test-beds in 2015. The main purpose of 5G test-beds is to help us better understand different 5G candidate technologies and come up with new 5G service use cases.

LR: What are the use cases that you find the most exciting for 5G?

SK: 5G will introduce many innovative 5G services, which fall into the following three types: 1) virtual experience, 2) massive IoT, and 3) mission-critical IoT. Customers will enjoy more immersive and virtual experience anytime and anywhere in a fully mobile and connected society. It is apparent that a lot more devices and "things" will become connected to enable massive IoT services to make our everyday lives more convenient. One of the differentiated values which we expect from 5G is the lower latency essential for mission-critical IoT services such as remote-controlled robots and connected cars.

Next page: Challenges, requirements, business models

LR: What do you see as the biggest challenges to the deployment of 5G?

SK: I see frequency spectrum and its global harmonization as one of key challenges for 5G commercialization. According to an ITU-R report, we would need several hundreds of bandwidth [sic] to meet market demands in the 5G era. However, it is obvious that there is not so much bandwidth available under 6GHz, therefore we believe that above-6GHz will play a key role in 5G radio access networks. Of course, it does not mean that we are not going to use sub-6GHz because this frequency is still very important to provide good cellular coverage. Finding globally harmonized frequency bands that meet 5G requirements both below and above 6GHz still remains challenging as different countries have their own regulations. We expect to resolve some of these issues in the upcoming WRC-15 led by ITU-R. Nevertheless, all 5G stakeholders need more discussion on globally harmonized frequency bands for 5G.

LR: How will 5G meet the diverse and challenging performance requirements (ie. 1,000-fold capacity gains, 10Gbit/s speeds, 1 ms latency and low power) all at the same time?

SK: Practically speaking, there will be no one-size-fits-all architectural solution to address the diverse 5G performance requirements all at the same time. On the other hand, not all services have every 5G service requirement. Therefore, we believe that the requirements will be met per service-basis using a divide-and-conquer approach. There will be many key enabling technologies of different costs available in different horizontal layers. In order to provide a vertical service in a cost-effective manner, it becomes important that the necessary technologies and resources from different horizontal layers will be systematically allocated, isolated and orchestrated.

LR: How will 5G change wireless business models for operators and consumers?

SK: 5G is often described as disruptive in nature. This is because 5G can potentially subsume other industries, given sufficient connectivity and bandwidth. The mobile broadband service to provide fast and reliable delivery of rich voice and immersive multimedia content (B2C) will obviously remain and continue to be one of the major business models. In addition, a significant growth in B2B & B2G [business-to-government] markets such as surveillance, health, smart-factories are expected, given the recent technology advances including machine-type-communication (MTC) and vehicle-to-anything (V2X) technologies.

LR: How should wireless operators think about investing in 5G when they are still trying to build out their 4G LTE networks?

SK: As we said earlier, we define 5G as 4G evolution together with new technologies. Therefore, we see that making smart investments on the 4G evolution technologies (such as NFV/SDN) would mean making partial investments on 5G.

Regardless of the stage a given operator is at in terms of building out 4G networks, the operator should have a plan to slowly embrace a paradigm shift in terms of infrastructure, operation and management. More specifically, the infrastructure will become more IT commercially-off-the-shelf (COTS) in nature, and the operations and management (O&M) will be increasingly automated.

— Sarah Thomas, , Editorial Operations Director, Light Reading